https://github.com/blankscreen-exe/data-mining-lab
Course Code: CS626, MCS Batch-2019 (Final Year) Evening
https://github.com/blankscreen-exe/data-mining-lab
data-visualization datamining datamining-algorithms datascience
Last synced: over 1 year ago
JSON representation
Course Code: CS626, MCS Batch-2019 (Final Year) Evening
- Host: GitHub
- URL: https://github.com/blankscreen-exe/data-mining-lab
- Owner: Blankscreen-exe
- Created: 2021-11-20T20:35:17.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2025-01-16T18:12:05.000Z (over 1 year ago)
- Last Synced: 2025-01-16T19:25:04.274Z (over 1 year ago)
- Topics: data-visualization, datamining, datamining-algorithms, datascience
- Language: Jupyter Notebook
- Homepage:
- Size: 18.4 MB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Data-Mining-Lab
Course Code: CS626
Course Supervisor: Dr. Tahseen Ahmed jilani
_Department of Computer Science_
_University of Karachi_
### Data Source:
You can find the data source on [Kaggle.com](https://www.kaggle.com/imakash3011/customer-personality-analysis?select=marketing_campaign.csv)
### Project Details:
Project details and description will be found in the documentation present in this repository
# Experiment
**Objectives:**
1. Predict the customer's response to marketting offers.
2. Perform Clustering Techniques to categorize the customer base.
------------------
#### Contents:
1. __Data OverView & Preprocessing__
- [Checking for Null Values](#Checking-for-Null-Values)
- [Checking Duplicate Values](#Checking-Duplicate-Values)
- [Looking at Uniques values](#Looking-at-Uniques-values)
- [Some More Tweaking](#Some-More-Tweaking)
- [Checking for Outliers](#Checking-for-Outliers)
- [Final Pre-Processsed Dataset](#Final-Pre-Processsed-Dataset)
2. __Exploratory Data Analysis__
- [Univariate Analysis](#Univariate-Analysis)
- [Bivariate Analysis](#Bivariate-Analysis)
- [Multivariate Analysis](#Multivariate-Analysis)
3. __Feature Selection & Dimensionality Reduction__
- [Preparing Sample for Prediction Models](#Preparing-Sample-for-Prediction-Models)
- [Feature Selection by Random Forest](#Feature-Selection-by-Random-Forest)
- [PCA Transformation](#PCA-Transformation)
4. __Supervised Predictions__
- [Preparing Data For Classification Models](#Preparing-Data-For-Classification-Models)
- [Logistic Regression](#Logistic-Regression)
- [Boosting Tree](#Boosting-Tree)
- [SVM](#SVM)
- [Neural Networks](#Neural-Networks)
- [Performance Comparison Among All 4 Models](#Performance-Comparison-Among-All-4-Models)
- [Final Model Performance](#Final-Model-Performance)
5. __Un-Supervised Predictions__
- [Feature Engineering and Clustering](#Feature-Engineering-and-Clustering)
- [K-Means](#K-Means)
- [Gaussian Mixture Model](#Gaussian-Mixture-Model)
6. __Summary__
- [1_ Customer-Related Summary](#1_-Customer-Related-Summary)
- [2_ Supervised Prediction Summary](#2_-Supervised-Prediction-Summary)
- [3_ Unsupervised Prediction Summary](#3_-Unsupervised-Prediction-Summary)
----
```python
#importing required modules
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import seaborn as sns
import scipy
from scipy.stats.mstats import winsorize
import random
import math
from tqdm import tqdm
from sklearn import preprocessing
from sklearn.metrics import matthews_corrcoef
from sklearn.metrics import accuracy_score
from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import KFold
from sklearn.linear_model import LogisticRegression
from sklearn.kernel_ridge import KernelRidge
from sklearn.ensemble import AdaBoostClassifier
from sklearn.metrics import fbeta_score, make_scorer
from sklearn import svm
import plotly.graph_objects as go
import plotly.express as px
import plotly.figure_factory as ff
%matplotlib inline
```
## 1. Data OverView & Preprocessing
```python
# Had to use '\t' separator, since all the rows were merged into one column cell.
Sample = pd.read_csv('marketing_campaign.csv',sep='\t')
Sample.head()
```
ID
Year_Birth
Education
Marital_Status
Income
Kidhome
Teenhome
Dt_Customer
Recency
MntWines
...
NumWebVisitsMonth
AcceptedCmp3
AcceptedCmp4
AcceptedCmp5
AcceptedCmp1
AcceptedCmp2
Complain
Z_CostContact
Z_Revenue
Response
0
5524
1957
Graduation
Single
58138.0
0
0
04-09-2012
58
635
...
7
0
0
0
0
0
0
3
11
1
1
2174
1954
Graduation
Single
46344.0
1
1
08-03-2014
38
11
...
5
0
0
0
0
0
0
3
11
0
2
4141
1965
Graduation
Together
71613.0
0
0
21-08-2013
26
426
...
4
0
0
0
0
0
0
3
11
0
3
6182
1984
Graduation
Together
26646.0
1
0
10-02-2014
26
11
...
6
0
0
0
0
0
0
3
11
0
4
5324
1981
PhD
Married
58293.0
1
0
19-01-2014
94
173
...
5
0
0
0
0
0
0
3
11
0
5 rows × 29 columns
```python
# Listing the columns' names
Sample.info()
```
RangeIndex: 2240 entries, 0 to 2239
Data columns (total 29 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 ID 2240 non-null int64
1 Year_Birth 2240 non-null int64
2 Education 2240 non-null object
3 Marital_Status 2240 non-null object
4 Income 2216 non-null float64
5 Kidhome 2240 non-null int64
6 Teenhome 2240 non-null int64
7 Dt_Customer 2240 non-null object
8 Recency 2240 non-null int64
9 MntWines 2240 non-null int64
10 MntFruits 2240 non-null int64
11 MntMeatProducts 2240 non-null int64
12 MntFishProducts 2240 non-null int64
13 MntSweetProducts 2240 non-null int64
14 MntGoldProds 2240 non-null int64
15 NumDealsPurchases 2240 non-null int64
16 NumWebPurchases 2240 non-null int64
17 NumCatalogPurchases 2240 non-null int64
18 NumStorePurchases 2240 non-null int64
19 NumWebVisitsMonth 2240 non-null int64
20 AcceptedCmp3 2240 non-null int64
21 AcceptedCmp4 2240 non-null int64
22 AcceptedCmp5 2240 non-null int64
23 AcceptedCmp1 2240 non-null int64
24 AcceptedCmp2 2240 non-null int64
25 Complain 2240 non-null int64
26 Z_CostContact 2240 non-null int64
27 Z_Revenue 2240 non-null int64
28 Response 2240 non-null int64
dtypes: float64(1), int64(25), object(3)
memory usage: 507.6+ KB
**COMMENT:** *we mostly have integer data types. Discrete methods might become necessary.*
### Checking for Null Values
```python
# Checking for Null Values
Sample.isna().sum()
```
ID 0
Year_Birth 0
Education 0
Marital_Status 0
Income 24
Kidhome 0
Teenhome 0
Dt_Customer 0
Recency 0
MntWines 0
MntFruits 0
MntMeatProducts 0
MntFishProducts 0
MntSweetProducts 0
MntGoldProds 0
NumDealsPurchases 0
NumWebPurchases 0
NumCatalogPurchases 0
NumStorePurchases 0
NumWebVisitsMonth 0
AcceptedCmp3 0
AcceptedCmp4 0
AcceptedCmp5 0
AcceptedCmp1 0
AcceptedCmp2 0
Complain 0
Z_CostContact 0
Z_Revenue 0
Response 0
dtype: int64
**COMMENT:** *Filling up the Null values with average (generic normalization method). Only Income column was affected, therefore we mostly have original data.*
```python
Sample['Income'].fillna(np.mean(Sample['Income']), inplace=True)
# converting currency type for better evaluation.
Sample['Income'] = Sample['Income'] / 1000
# Just for confirmation of Null value removal
for i in Sample.isna().sum():
if i != 0:
print(i)
else:
print("Filling Null spaces Successful!")
```
Filling Null spaces Successful!
### Checking Duplicate Values
```python
Sample.duplicated().sum()
```
0
**COMMENT:** *No duplicates were found*
### Looking at Uniques values
```python
pd.DataFrame(Sample.nunique()).sort_values(0).rename( {0: 'Unique Values'}, axis=1)
```
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
Unique Values
Z_Revenue
1
Z_CostContact
1
Response
2
AcceptedCmp3
2
AcceptedCmp4
2
AcceptedCmp2
2
Complain
2
AcceptedCmp1
2
AcceptedCmp5
2
Kidhome
3
Teenhome
3
Education
5
Marital_Status
8
NumCatalogPurchases
14
NumStorePurchases
14
NumDealsPurchases
15
NumWebPurchases
15
NumWebVisitsMonth
16
Year_Birth
59
Recency
100
MntFruits
158
MntSweetProducts
177
MntFishProducts
182
MntGoldProds
213
MntMeatProducts
558
Dt_Customer
663
MntWines
776
Income
1975
ID
2240
**COMMENT:**
- *Z_Revenue and Z_CostContact have only single values for all entries, therefore we will remove those columns*
- *Other entries are good to go. Except for ID and Dt_Customer, they won't be necessary in our analysis*
```python
# Dropping the "Z_Revenue" and "Z_CostContact" columns
Sample.drop(['Z_CostContact', 'Z_Revenue'], axis=1, inplace=True)
# Also dropping "ID" and "Dt_Customer", since they are not useful in our processing.
Sample.drop(['ID', 'Dt_Customer'], axis=1, inplace=True)
```
### Some More Tweaking
```python
# Converting "Year_Birth" to "Age"
Sample['Age'] = 2021 - Sample.Year_Birth.to_numpy()
Sample.drop('Year_Birth', axis=1, inplace=True)
```
```python
Sample['Marital_Status'].value_counts()
```
Married 864
Together 580
Single 480
Divorced 232
Widow 77
Alone 3
YOLO 2
Absurd 2
Name: Marital_Status, dtype: int64
**COMMENT:** *This categorization is not good, since many factors suggest that **Alone, YOLO, Absurd** should be regrouped into **Single***
```python
# Re-grouping "Marital_Status" Column
Sample['Marital_Status'] = Sample['Marital_Status'].replace(['Alone','YOLO','Absurd'],'Single')
```
```python
Sample['Education'].value_counts()
```
Graduation 1127
PhD 486
Master 370
2n Cycle 203
Basic 54
Name: Education, dtype: int64
**COMMENT:** *This categorization shows too many unnecessary groups. For example, **2n-Cycle** is equivalent to **Masters***
```python
Sample['Education'].replace(['2n Cycle', 'Graduation'], ['Master', 'Bachelor'], inplace=True)
```
### Checking for Outliers
```python
# using box plot to visualize outliers in numerial fields of the dataset
n = Sample.select_dtypes(include=np.number).columns.tolist()
bins=10
j=1
fig = plt.figure(figsize = (20, 30))
for i in n:
plt.subplot(7,4,j)
plt.boxplot(Sample[i])
j=j+1
plt.xlabel(i)
plt.show()
```
**COMMENT:** *Since we are sure about outliers in "Age" and "Income" pose unnecessary plots, therefore we can remove them. As for other fields, we are not sure, therefore we wil let them be.*
```python
Sample.drop(Sample[(Sample['Income']>200)|(Sample['Age']>100)].index,inplace=True)
```
### Final Pre-Processsed Dataset
```python
Sample.head()
```
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
Education
Marital_Status
Income
Kidhome
Teenhome
Recency
MntWines
MntFruits
MntMeatProducts
MntFishProducts
...
NumStorePurchases
NumWebVisitsMonth
AcceptedCmp3
AcceptedCmp4
AcceptedCmp5
AcceptedCmp1
AcceptedCmp2
Complain
Response
Age
0
Bachelor
Single
58.138
0
0
58
635
88
546
172
...
4
7
0
0
0
0
0
0
1
64
1
Bachelor
Single
46.344
1
1
38
11
1
6
2
...
2
5
0
0
0
0
0
0
0
67
2
Bachelor
Together
71.613
0
0
26
426
49
127
111
...
10
4
0
0
0
0
0
0
0
56
3
Bachelor
Together
26.646
1
0
26
11
4
20
10
...
4
6
0
0
0
0
0
0
0
37
4
PhD
Married
58.293
1
0
94
173
43
118
46
...
6
5
0
0
0
0
0
0
0
40
5 rows × 25 columns
-----
## 2. Exploratory Data Analysis
### Univariate Analysis
We are mainly looking at each variable separately in order to derive insights from it.
```python
# Marital Status and Education Levels in a pie chart
c_count = Sample.groupby("Marital_Status").count()['Age']
label = Sample.groupby('Marital_Status').count()['Age'].index
fig, ax = plt.subplots(1, 2, figsize = (10, 12))
ax[0].pie(c_count, labels=label, colors=sns.color_palette('pastel')[0:5], autopct='%1.2f%%',radius=2,explode=[0.07,0.07,0.07,0.07,0.07])
ax[0].set_title('Maritial Status', y=-0.6)
c_count = Sample.groupby("Education").count()['Age']
label = Sample.groupby('Education').count()['Age'].index
ax[1].pie(c_count, labels=label, colors=sns.color_palette('pastel')[0:5], autopct='%1.2f%%',radius=2,explode=[0.07,0.07,0.07,0.07])
ax[1].set_title('Education Level', y=-0.6)
plt.subplots_adjust(wspace = 1.5, hspace =0)
plt.show()
```
**COMMENT:** *We have more Married Couples or Bachelors Degree holders as our customers. Approx. 39% and 50% respectively.*
```python
# Age
plt.figure(figsize=(20, 6))
plt.title('Age distribution')
ax = sns.histplot(Sample['Age'].sort_values(), bins=56, color='skyblue')
sns.rugplot(data=Sample['Age'], height=.05)
plt.xticks(np.linspace(Sample['Age'].min(), Sample['Age'].max(), 56, dtype=int, endpoint = True))
plt.show()
```
**COMMENT:** *Most of our customers are people among the age of 45-60 years*
```python
# Home with kids and Home with Teens
plt.figure(figsize=(15,5))
plt.subplot(121)
sns.histplot(data=Sample, x='Kidhome', stat="count", discrete=True)
plt.xticks([0, 1, 2])
plt.subplot(122)
sns.histplot(data=Sample, x='Teenhome', stat="count",discrete=True)
plt.xticks([0, 1, 2])
plt.show()
```
**COMMENT:** *A very small group of people have about either 2 kids or 2 Teens in their home. Other types of customers are significantly large.*
```python
# Response dist. and Income Range
plt.figure(figsize=(15,5))
plt.subplot(121)
sns.histplot(data=Sample, x='Response', stat="count", discrete=True)
plt.xticks([0, 1])
plt.subplot(122)
# Income Range
sns.kdeplot(data=Sample, x="Income", shade=True, log_scale=True)
plt.show()
```
**COMMENT:**
- *This shows a very bad situation. The last campaign was **mostly rejected***
- *Income of most people range from 10,000 to 100,000 units*
-----
### Bivariate Analysis
Here we are performing a comparative analysis of how the variables relate together to each other
```python
# Let's see if "Education" level have any impact on the "Responses" we got.
plt.figure(figsize=(15,5))
plt.subplot(121)
sns.histplot(data=Sample, x="Education", hue="Response", multiple="stack", stat="count",color='skyblue')
# Now we will look for relations among "Marital Status" and "Responses"
plt.subplot(122)
sns.histplot(data=Sample, x="Marital_Status", hue="Response",stat="count", multiple="stack", color='skyblue')
plt.show()
```
**COMMENT:**
- *Left figure; people having low education levels were not in the least interested in this campaign.*
- *Right figure; Single people were more likely to accept our offers*
```python
# Let's see if "Homw with kids" level have any impact on the "Responses" we got.
plt.figure(figsize=(15,5))
plt.subplot(121)
sns.histplot(data=Sample, x="Kidhome", hue="Response", multiple="stack", stat="count", discrete=True)
plt.xticks([0, 1, 2])
# Now we will look for relations among "Home with teens" and "Responses"
plt.subplot(122)
sns.histplot(data=Sample, x="Teenhome", hue="Response", multiple="stack", stat="count", discrete=True)
plt.xticks([0, 1, 2])
plt.show()
```
**COMMENT:** *People with no kids or teens at home were more interested in our campaign*
```python
# Let's see if "Income" has any impact on the "Responses", "Marital Status", "Education", "Kid Home".
plt.figure(figsize=(15,10))
plt.subplot(221)
sns.kdeplot(
data=Sample, x="Income", hue="Response", log_scale= True,
fill=True, common_norm=False,
alpha=.5, linewidth=0,
)
plt.gca().axes.get_yaxis().set_visible(False) # Set y invisible
plt.xlabel('Income')
# segment by Marital_Status
plt.subplot(222)
sns.kdeplot(
data=Sample, x="Income", hue="Marital_Status", log_scale= True,
fill=True, common_norm=False, palette="crest",
alpha=.5, linewidth=0,
)
plt.gca().axes.get_yaxis().set_visible(False)
# segment by Education
plt.subplot(223)
sns.kdeplot(
data=Sample, x="Income", hue="Education", log_scale= True,
fill=True, common_norm=False, palette="crest",
alpha=.5, linewidth=0,
)
plt.gca().axes.get_yaxis().set_visible(False)
# segment by Kidhome
plt.subplot(224)
sns.kdeplot(
data=Sample, x="Income", hue="Kidhome", log_scale= True,
fill=True, common_norm=False, palette="crest",
alpha=.5, linewidth=0,
)
plt.gca().axes.get_yaxis().set_visible(False)
```
**COMMENT:**
- *People with more Income are more likely to accept our offers. We can say that people with income 14-15,000 or less don't seem that much interested.*
- *Different Marital Status does not seem to be the cause of positive or negative response to our marketing campaign.*
- *People having lower level of education have less income. Those Bachelors, Masters or Ph.D degrees do not have clear difference between their incomes.*
- *Those who do not have kids at home have higher income.*
----
### Multivariate Analysis
Here we are going to compare multiple variables to see if they any relations.
```python
# Re-stating variable types
pd.DataFrame(Sample.nunique()).sort_values(0).rename( {0: 'Unique Values'}, axis=1)
```
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
Unique Values
Complain
2
AcceptedCmp2
2
AcceptedCmp1
2
AcceptedCmp5
2
AcceptedCmp4
2
AcceptedCmp3
2
Response
2
Kidhome
3
Teenhome
3
Education
4
Marital_Status
5
NumStorePurchases
14
NumCatalogPurchases
14
NumDealsPurchases
15
NumWebPurchases
15
NumWebVisitsMonth
16
Age
56
Recency
100
MntFruits
158
MntSweetProducts
177
MntFishProducts
182
MntGoldProds
213
MntMeatProducts
557
MntWines
775
Income
1971
```python
# looking for linear trends between numerical values
NUMERICAL_FEATURES = ['Age', 'Recency', 'MntFruits',
'MntSweetProducts', 'MntFishProducts', 'MntGoldProds',
'MntMeatProducts', 'MntWines', 'Income']
sns.pairplot(data=Sample[NUMERICAL_FEATURES],
kind='scatter', plot_kws={'alpha':0.4})
plt.show()
```
```python
# Heat map among all numerical variables
cor = Sample.corr()
plt.figure(figsize = (27,26))
sns.heatmap(cor, annot = True, cmap = 'coolwarm')
plt.show()
```
**COMMENT:** *We found no linear relationships between the numerical variables*
----
----
## 3. Feature Selection & Dimensionality Reduction
### Preparing Sample for Prediction Models
```python
# Converting Education with scale numerals
df_cluster = Sample.copy() # saving for clustering
Sample['Education'] = Sample['Education'].replace(['Basic'], 0)
Sample['Education'] = Sample['Education'].replace(['Bachelor'], 1)
Sample['Education'] = Sample['Education'].replace(['Master'], 2)
Sample['Education'] = Sample['Education'].replace(['PhD'], 3)
# Change Marital_Status to dummies
Sample = pd.get_dummies(Sample)
Sample.info()
```
Int64Index: 2236 entries, 0 to 2239
Data columns (total 29 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 Education 2236 non-null int64
1 Income 2236 non-null float64
2 Kidhome 2236 non-null int64
3 Teenhome 2236 non-null int64
4 Recency 2236 non-null int64
5 MntWines 2236 non-null int64
6 MntFruits 2236 non-null int64
7 MntMeatProducts 2236 non-null int64
8 MntFishProducts 2236 non-null int64
9 MntSweetProducts 2236 non-null int64
10 MntGoldProds 2236 non-null int64
11 NumDealsPurchases 2236 non-null int64
12 NumWebPurchases 2236 non-null int64
13 NumCatalogPurchases 2236 non-null int64
14 NumStorePurchases 2236 non-null int64
15 NumWebVisitsMonth 2236 non-null int64
16 AcceptedCmp3 2236 non-null int64
17 AcceptedCmp4 2236 non-null int64
18 AcceptedCmp5 2236 non-null int64
19 AcceptedCmp1 2236 non-null int64
20 AcceptedCmp2 2236 non-null int64
21 Complain 2236 non-null int64
22 Response 2236 non-null int64
23 Age 2236 non-null int64
24 Marital_Status_Divorced 2236 non-null uint8
25 Marital_Status_Married 2236 non-null uint8
26 Marital_Status_Single 2236 non-null uint8
27 Marital_Status_Together 2236 non-null uint8
28 Marital_Status_Widow 2236 non-null uint8
dtypes: float64(1), int64(23), uint8(5)
memory usage: 447.6 KB
---
### Feature Selection by Random Forest
```python
# Using Random Forest to gain an insight on Feature Importance
clf = RandomForestClassifier()
clf.fit(Sample.drop('Response', axis=1), Sample['Response'])
plt.style.use('seaborn-whitegrid')
importance = clf.feature_importances_
importance = pd.DataFrame(importance, index=Sample.drop('Response', axis=1).columns, columns=["Importance"])
importance.sort_values(by='Importance', ascending=True).plot(kind='barh', figsize=(20,len(importance)/2));
```
```python
# Choosing Features (Keep 90% Importance ratio)
feature_nums = 18
ascend_import = importance.sort_values(by='Importance', ascending=False)
all_info = ascend_import['Importance'].iloc[:feature_nums].sum()
all_choose_features = list(ascend_import.iloc[:feature_nums].index)
for i in range(len(all_choose_features)):
if i%2==0:
print(all_choose_features[i]+' ; ', end="")
else:
print(all_choose_features[i])
print('\n\nImportance Raio: ', all_info)
```
Recency ; MntWines
Income ; MntMeatProducts
MntGoldProds ; Age
AcceptedCmp3 ; MntSweetProducts
AcceptedCmp5 ; MntFruits
MntFishProducts ; NumWebVisitsMonth
NumStorePurchases ; NumCatalogPurchases
NumWebPurchases ; NumDealsPurchases
AcceptedCmp1 ; Education
Importance Raio: 0.9097738142686715
---
### PCA Transformation
```python
from sklearn.decomposition import PCA
# Calculating PCA for both datasets, and graphing the Variance for each feature, per dataset
std_scale = preprocessing.StandardScaler().fit(Sample.drop('Response', axis=1))
X = std_scale.transform(Sample.drop('Response', axis=1))
pca1 = PCA(0.90, whiten=True) # Keep 90% information
fit1 = pca1.fit(X)
# Graphing the variance per feature
plt.style.use('seaborn-whitegrid')
plt.figure(figsize=(25,7))
plt.xlabel('PCA Feature')
plt.ylabel('Variance')
plt.title('PCA for Whole Dataset')
plt.bar(range(0, fit1.explained_variance_ratio_.size), fit1.explained_variance_ratio_);
# Get pca transformed data
pca_data = pca1.transform(X)
pca_data = np.array(pca_data)
print('PCA data shape: ', pca_data.shape)
```
PCA data shape: (2236, 19)
```python
# Stores performances of models
perf_df_lst = [None, None, None]
```
```python
# Data sourse encoding
# 0 --> Raw Data
# 1 --> Selected Features
# 2 --> PCA
# Choose dataset
dataset_num = 1
# 0: Raw Data; 1: Feature Selection Data; 2: PCA Data
all_datasets = [Sample.drop('Response', axis=1).values, Sample[all_choose_features].values, pca_data]
# Choose data
final_data = all_datasets[dataset_num]
```
---
---
## 4. Supervised Predictions
### Preparing Data For Classification Models
```python
# Split the dataset
from imblearn.over_sampling import SMOTE
from collections import Counter
x_train = final_data[:2000]
y_train = Sample['Response'].values[:2000]
x_test = final_data[2000:]
y_test = Sample['Response'].values[2000:]
print('Train: ', len(x_train))
print('Test: ', len(x_test))
print('N/P Sample: ', Counter(y_train))
# SMOTE Samples
sm = SMOTE(random_state=2)
x_train, y_train = sm.fit_resample(x_train, y_train.ravel())
print('Train: ', len(x_train))
print('Test: ', len(x_test))
print('N/P Sample: ', Counter(y_train))
# MCC scorer function
mcc_scorer = make_scorer(matthews_corrcoef)
```
Train: 2000
Test: 236
N/P Sample: Counter({0: 1698, 1: 302})
Train: 3396
Test: 236
N/P Sample: Counter({1: 1698, 0: 1698})
---
### Logistic Regression
```python
LR = LogisticRegression()
# K-Fold Validation
kfold = 8
# ACC Score
LR_cv_results_acc = cross_val_score(LR, x_train, y_train, cv=kfold, scoring='accuracy')
msg = "%s k-fold ACC: %f (%f)" % ('LR', LR_cv_results_acc.mean(), LR_cv_results_acc.std())
print(msg)
# MCC Score
LR_cv_results_mcc = cross_val_score(LR, x_train, y_train, cv=kfold, scoring=mcc_scorer)
msg = "%s k-fold MCC: %f (%f)" % ('LR', LR_cv_results_mcc.mean(), LR_cv_results_mcc.std())
print(msg)
```
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
LR k-fold ACC: 0.750893 (0.017630)
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
LR k-fold MCC: 0.502319 (0.035211)
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
C:\Users\D4Rk_C4K3\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
Increase the number of iterations (max_iter) or scale the data as shown in:
https://scikit-learn.org/stable/modules/preprocessing.html
Please also refer to the documentation for alternative solver options:
https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
n_iter_i = _check_optimize_result(
OUTPUT>>LR k-fold ACC: 0.749420 (0.015291)
OUTPUT>>LR k-fold MCC: 0.499386 (0.030593)
---
### Boosting Tree
```python
# Validation for Boosting Tree
clf = AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=2), n_estimators=160, random_state=1)
# ACC Score
BT_cv_results_acc = cross_val_score(clf, x_train, y_train, cv=kfold, scoring='accuracy')
msg = "k-fold ACC: %f (%f)" % (BT_cv_results_acc.mean(), BT_cv_results_acc.std())
print(msg)
# MCC Score
BT_cv_results_mcc = cross_val_score(clf, x_train, y_train, cv=kfold, scoring=mcc_scorer)
msg = "k-fold MCC: %f (%f)" % (BT_cv_results_mcc.mean(), BT_cv_results_mcc.std())
print(msg)
```
k-fold ACC: 0.903476 (0.098889)
k-fold MCC: 0.819825 (0.173745)
---
### SVM
```python
SVM=svm.SVC(kernel = 'rbf', C = 10, gamma = 0.01)
# ACC Score
svm_cv_results_acc = cross_val_score(SVM, x_train, y_train, cv=kfold, scoring='accuracy')
msg = "k-fold ACC: %f (%f)" % (svm_cv_results_acc.mean(), svm_cv_results_acc.std())
print(msg)
# MCC Score
svm_cv_results_mcc = cross_val_score(SVM, x_train, y_train, cv=kfold, scoring=mcc_scorer)
msg = "k-fold MCC: %f (%f)" % (svm_cv_results_mcc.mean(), svm_cv_results_mcc.std())
print(msg)
```
k-fold ACC: 0.802105 (0.023620)
k-fold MCC: 0.644576 (0.039894)
---
### Neural Networks
```python
# import modules
from keras import models
from keras import layers
x_nn = x_train
y_nn = y_train
```
```python
# Define the K-fold Cross Validator
kfold_obj = KFold(n_splits=kfold, shuffle=True)
epochs = 150
batch_size = 64
acc_per_fold = []
mcc_per_fold = []
# K-fold Cross Validation model evaluation
fold_no = 1
for train, test in kfold_obj.split(x_nn, y_nn):
model = models.Sequential()
# Only use shallow Neural Network
model.add(layers.Dense(32, activation='relu'))
model.add(layers.Dense(1, activation='sigmoid'))
model.compile(optimizer='rmsprop', loss='binary_crossentropy', metrics=['accuracy'])
# Generate a print
print('------------------------------------------------------------------------')
print(f'Training for fold {fold_no} ...')
# Fit data to model
history = model.fit(x_nn[train], y_nn[train].reshape(-1, 1),
batch_size=64,
epochs=100)
# Generate generalization metrics
predictions = model.predict(x_nn[test])
for i in range(len(predictions)):
if predictions[i] >=0.5:
predictions[i] = 1
else:
predictions[i] = 0
acc_per_fold.append(accuracy_score(predictions, y_nn[test]))
mcc_per_fold.append(matthews_corrcoef(predictions, y_nn[test]))
# Increase fold number
fold_no = fold_no + 1
```
------------------------------------------------------------------------
Training for fold 1 ...
Epoch 1/100
47/47 [==============================] - 1s 1ms/step - loss: 14.7443 - accuracy: 0.5150
Epoch 2/100
47/47 [==============================] - 0s 1ms/step - loss: 5.0344 - accuracy: 0.5914
Epoch 3/100
47/47 [==============================] - 0s 1ms/step - loss: 3.0739 - accuracy: 0.6473
Epoch 4/100
47/47 [==============================] - 0s 1ms/step - loss: 2.2507 - accuracy: 0.6695
Epoch 5/100
47/47 [==============================] - 0s 1ms/step - loss: 1.8620 - accuracy: 0.6846
Epoch 6/100
47/47 [==============================] - 0s 1ms/step - loss: 1.5295 - accuracy: 0.6974
Epoch 7/100
47/47 [==============================] - 0s 2ms/step - loss: 1.5116 - accuracy: 0.6860
Epoch 8/100
47/47 [==============================] - 0s 1ms/step - loss: 1.3697 - accuracy: 0.7031
Epoch 9/100
47/47 [==============================] - 0s 1ms/step - loss: 1.2780 - accuracy: 0.7072
Epoch 10/100
47/47 [==============================] - 0s 1ms/step - loss: 1.2435 - accuracy: 0.7196
Epoch 11/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1951 - accuracy: 0.7183
Epoch 12/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1365 - accuracy: 0.7264
Epoch 13/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1171 - accuracy: 0.7210
Epoch 14/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1885 - accuracy: 0.7280
Epoch 15/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1463 - accuracy: 0.7189
Epoch 16/100
47/47 [==============================] - 0s 1ms/step - loss: 1.0773 - accuracy: 0.7304
Epoch 17/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9812 - accuracy: 0.7408
Epoch 18/100
47/47 [==============================] - 0s 999us/step - loss: 1.0379 - accuracy: 0.7341
Epoch 19/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9865 - accuracy: 0.7513
Epoch 20/100
47/47 [==============================] - 0s 978us/step - loss: 0.9898 - accuracy: 0.7368
Epoch 21/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9581 - accuracy: 0.7489
Epoch 22/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9547 - accuracy: 0.7492
Epoch 23/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9477 - accuracy: 0.7573
Epoch 24/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9834 - accuracy: 0.7391
Epoch 25/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9715 - accuracy: 0.7472
Epoch 26/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9398 - accuracy: 0.7452
Epoch 27/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8465 - accuracy: 0.7610
Epoch 28/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9573 - accuracy: 0.7381
Epoch 29/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8786 - accuracy: 0.7694
Epoch 30/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8935 - accuracy: 0.7540
Epoch 31/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8247 - accuracy: 0.7711
Epoch 32/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8897 - accuracy: 0.7580
Epoch 33/100
47/47 [==============================] - 0s 978us/step - loss: 0.8743 - accuracy: 0.7556
Epoch 34/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8854 - accuracy: 0.7529
Epoch 35/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8520 - accuracy: 0.7684
Epoch 36/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8483 - accuracy: 0.7604
Epoch 37/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8179 - accuracy: 0.7765
Epoch 38/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8731 - accuracy: 0.7688
Epoch 39/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7984 - accuracy: 0.7698
Epoch 40/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8754 - accuracy: 0.7614
Epoch 41/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7177 - accuracy: 0.7876
Epoch 42/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8563 - accuracy: 0.7600
Epoch 43/100
47/47 [==============================] - 0s 999us/step - loss: 0.8097 - accuracy: 0.7708
Epoch 44/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7812 - accuracy: 0.7829
Epoch 45/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8258 - accuracy: 0.7691
Epoch 46/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8157 - accuracy: 0.7772
Epoch 47/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7862 - accuracy: 0.7765
Epoch 48/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8348 - accuracy: 0.7738
Epoch 49/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7472 - accuracy: 0.7826
Epoch 50/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7795 - accuracy: 0.7839
Epoch 51/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7589 - accuracy: 0.7795
Epoch 52/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7744 - accuracy: 0.7799
Epoch 53/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7907 - accuracy: 0.7762
Epoch 54/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7384 - accuracy: 0.7960
Epoch 55/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7836 - accuracy: 0.7859
Epoch 56/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7453 - accuracy: 0.7782
Epoch 57/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7499 - accuracy: 0.7893
Epoch 58/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7824 - accuracy: 0.7856
Epoch 59/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7963 - accuracy: 0.7789
Epoch 60/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7697 - accuracy: 0.7856
Epoch 61/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7260 - accuracy: 0.7896
Epoch 62/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7739 - accuracy: 0.7883
Epoch 63/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6969 - accuracy: 0.7886
Epoch 64/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7688 - accuracy: 0.7805
Epoch 65/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7668 - accuracy: 0.7846
Epoch 66/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7163 - accuracy: 0.7910
Epoch 67/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7695 - accuracy: 0.7822
Epoch 68/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7634 - accuracy: 0.7819
Epoch 69/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6730 - accuracy: 0.8011
Epoch 70/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7333 - accuracy: 0.7923
Epoch 71/100
47/47 [==============================] - 0s 2ms/step - loss: 0.7284 - accuracy: 0.7954
Epoch 72/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7717 - accuracy: 0.7859
Epoch 73/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7246 - accuracy: 0.8004
Epoch 74/100
47/47 [==============================] - 0s 2ms/step - loss: 0.7051 - accuracy: 0.8021
Epoch 75/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7574 - accuracy: 0.7900
Epoch 76/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7083 - accuracy: 0.7991
Epoch 77/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7286 - accuracy: 0.8014
Epoch 78/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6732 - accuracy: 0.8034
Epoch 79/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6969 - accuracy: 0.8075
Epoch 80/100
47/47 [==============================] - 0s 2ms/step - loss: 0.7141 - accuracy: 0.7997
Epoch 81/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6954 - accuracy: 0.8075
Epoch 82/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7570 - accuracy: 0.7923
Epoch 83/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6240 - accuracy: 0.8078
Epoch 84/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7433 - accuracy: 0.8095
Epoch 85/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7100 - accuracy: 0.8021
Epoch 86/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6757 - accuracy: 0.8061
Epoch 87/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7369 - accuracy: 0.7984
Epoch 88/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6973 - accuracy: 0.8018
Epoch 89/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6682 - accuracy: 0.8075
Epoch 90/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6953 - accuracy: 0.8092
Epoch 91/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6747 - accuracy: 0.8129
Epoch 92/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6476 - accuracy: 0.8112
Epoch 93/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6639 - accuracy: 0.8102
Epoch 94/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7606 - accuracy: 0.7883
Epoch 95/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6458 - accuracy: 0.8031
Epoch 96/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6512 - accuracy: 0.8055
Epoch 97/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6951 - accuracy: 0.8125
Epoch 98/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6170 - accuracy: 0.8162
Epoch 99/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6618 - accuracy: 0.8189
Epoch 100/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6749 - accuracy: 0.8018
------------------------------------------------------------------------
Training for fold 2 ...
Epoch 1/100
47/47 [==============================] - 1s 978us/step - loss: 4.3806 - accuracy: 0.5540
Epoch 2/100
47/47 [==============================] - 0s 956us/step - loss: 2.1046 - accuracy: 0.6318
Epoch 3/100
47/47 [==============================] - 0s 1ms/step - loss: 1.4318 - accuracy: 0.6479
Epoch 4/100
47/47 [==============================] - 0s 1ms/step - loss: 1.1510 - accuracy: 0.6624
Epoch 5/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9356 - accuracy: 0.6839
Epoch 6/100
47/47 [==============================] - 0s 999us/step - loss: 0.9172 - accuracy: 0.6951
Epoch 7/100
47/47 [==============================] - 0s 999us/step - loss: 0.8125 - accuracy: 0.7062
Epoch 8/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7715 - accuracy: 0.7223
Epoch 9/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7082 - accuracy: 0.7280
Epoch 10/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6704 - accuracy: 0.7334
Epoch 11/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6859 - accuracy: 0.7385
Epoch 12/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6281 - accuracy: 0.7395
Epoch 13/100
47/47 [==============================] - ETA: 0s - loss: 0.6104 - accuracy: 0.75 - 0s 1ms/step - loss: 0.6137 - accuracy: 0.7556
Epoch 14/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5945 - accuracy: 0.7540
Epoch 15/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5890 - accuracy: 0.7694
Epoch 16/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6035 - accuracy: 0.7684
Epoch 17/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5805 - accuracy: 0.7684
Epoch 18/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5599 - accuracy: 0.7752
Epoch 19/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5796 - accuracy: 0.7644
Epoch 20/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5541 - accuracy: 0.7785
Epoch 21/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5578 - accuracy: 0.7809
Epoch 22/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5530 - accuracy: 0.7782
Epoch 23/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5274 - accuracy: 0.7856
Epoch 24/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5028 - accuracy: 0.7906
Epoch 25/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5357 - accuracy: 0.7893
Epoch 26/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5084 - accuracy: 0.7977
Epoch 27/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5311 - accuracy: 0.7883
Epoch 28/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5142 - accuracy: 0.8011
Epoch 29/100
47/47 [==============================] - 0s 934us/step - loss: 0.5258 - accuracy: 0.7964
Epoch 30/100
47/47 [==============================] - 0s 913us/step - loss: 0.5024 - accuracy: 0.7970
Epoch 31/100
47/47 [==============================] - 0s 999us/step - loss: 0.4893 - accuracy: 0.8034
Epoch 32/100
47/47 [==============================] - 0s 956us/step - loss: 0.4985 - accuracy: 0.8014
Epoch 33/100
47/47 [==============================] - 0s 956us/step - loss: 0.5052 - accuracy: 0.7893
Epoch 34/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4715 - accuracy: 0.8068
Epoch 35/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4602 - accuracy: 0.8135
Epoch 36/100
47/47 [==============================] - 0s 934us/step - loss: 0.5203 - accuracy: 0.8004
Epoch 37/100
47/47 [==============================] - 0s 999us/step - loss: 0.5298 - accuracy: 0.8007
Epoch 38/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4733 - accuracy: 0.8075
Epoch 39/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4757 - accuracy: 0.8092
Epoch 40/100
47/47 [==============================] - 0s 978us/step - loss: 0.4918 - accuracy: 0.8105
Epoch 41/100
47/47 [==============================] - 0s 978us/step - loss: 0.4490 - accuracy: 0.8182
Epoch 42/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4816 - accuracy: 0.8095
Epoch 43/100
47/47 [==============================] - 0s 978us/step - loss: 0.4687 - accuracy: 0.8199
Epoch 44/100
47/47 [==============================] - 0s 912us/step - loss: 0.5016 - accuracy: 0.8048
Epoch 45/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4398 - accuracy: 0.8236
Epoch 46/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4694 - accuracy: 0.8092
Epoch 47/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4966 - accuracy: 0.8004
Epoch 48/100
47/47 [==============================] - 0s 978us/step - loss: 0.4511 - accuracy: 0.8230
Epoch 49/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4907 - accuracy: 0.8142
Epoch 50/100
47/47 [==============================] - 0s 978us/step - loss: 0.4565 - accuracy: 0.8152
Epoch 51/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4746 - accuracy: 0.8182
Epoch 52/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4593 - accuracy: 0.8176
Epoch 53/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4497 - accuracy: 0.8270
Epoch 54/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4462 - accuracy: 0.8250
Epoch 55/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4431 - accuracy: 0.8246
Epoch 56/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4476 - accuracy: 0.8216
Epoch 57/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4246 - accuracy: 0.8189
Epoch 58/100
47/47 [==============================] - 0s 934us/step - loss: 0.4275 - accuracy: 0.8314
Epoch 59/100
47/47 [==============================] - 0s 956us/step - loss: 0.4543 - accuracy: 0.8240
Epoch 60/100
47/47 [==============================] - 0s 999us/step - loss: 0.4713 - accuracy: 0.8061
Epoch 61/100
47/47 [==============================] - 0s 956us/step - loss: 0.4487 - accuracy: 0.8294
Epoch 62/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4507 - accuracy: 0.8182
Epoch 63/100
47/47 [==============================] - 0s 999us/step - loss: 0.4336 - accuracy: 0.8223
Epoch 64/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4404 - accuracy: 0.8216
Epoch 65/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4347 - accuracy: 0.8230
Epoch 66/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4416 - accuracy: 0.8246
Epoch 67/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4514 - accuracy: 0.8216
Epoch 68/100
47/47 [==============================] - 0s 999us/step - loss: 0.4281 - accuracy: 0.8314
Epoch 69/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4531 - accuracy: 0.8223
Epoch 70/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4152 - accuracy: 0.8324
Epoch 71/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4336 - accuracy: 0.8324
Epoch 72/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4387 - accuracy: 0.8327
Epoch 73/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4531 - accuracy: 0.8263
Epoch 74/100
47/47 [==============================] - 0s 956us/step - loss: 0.4300 - accuracy: 0.8371
Epoch 75/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4156 - accuracy: 0.8337
Epoch 76/100
47/47 [==============================] - 0s 978us/step - loss: 0.4268 - accuracy: 0.8324
Epoch 77/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4462 - accuracy: 0.8223
Epoch 78/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4412 - accuracy: 0.8243
Epoch 79/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4042 - accuracy: 0.8442
Epoch 80/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4363 - accuracy: 0.8294
Epoch 81/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4279 - accuracy: 0.8277
Epoch 82/100
47/47 [==============================] - 0s 934us/step - loss: 0.4178 - accuracy: 0.8485
Epoch 83/100
47/47 [==============================] - 0s 1000us/step - loss: 0.4161 - accuracy: 0.8337
Epoch 84/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4493 - accuracy: 0.8304
Epoch 85/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4169 - accuracy: 0.8260
Epoch 86/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4083 - accuracy: 0.8421
Epoch 87/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4303 - accuracy: 0.8458
Epoch 88/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4408 - accuracy: 0.8320
Epoch 89/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4176 - accuracy: 0.8384
Epoch 90/100
47/47 [==============================] - 0s 1ms/step - loss: 0.3922 - accuracy: 0.8469
Epoch 91/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4155 - accuracy: 0.8428
Epoch 92/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4061 - accuracy: 0.8388
Epoch 93/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4195 - accuracy: 0.8445
Epoch 94/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4073 - accuracy: 0.8435
Epoch 95/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4115 - accuracy: 0.8398
Epoch 96/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4461 - accuracy: 0.8280
Epoch 97/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4135 - accuracy: 0.8371
Epoch 98/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4219 - accuracy: 0.8411
Epoch 99/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4153 - accuracy: 0.8374
Epoch 100/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4134 - accuracy: 0.8374
------------------------------------------------------------------------
Training for fold 3 ...
Epoch 1/100
47/47 [==============================] - 1s 1ms/step - loss: 5.8383 - accuracy: 0.5523
Epoch 2/100
47/47 [==============================] - 0s 1ms/step - loss: 2.3424 - accuracy: 0.6341
Epoch 3/100
47/47 [==============================] - 0s 1ms/step - loss: 1.5177 - accuracy: 0.6459
Epoch 4/100
47/47 [==============================] - 0s 1ms/step - loss: 1.2274 - accuracy: 0.6590
Epoch 5/100
47/47 [==============================] - 0s 978us/step - loss: 1.1577 - accuracy: 0.6819
Epoch 6/100
47/47 [==============================] - 0s 1ms/step - loss: 1.0207 - accuracy: 0.6967
Epoch 7/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9979 - accuracy: 0.6998
Epoch 8/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9120 - accuracy: 0.7126
Epoch 9/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8612 - accuracy: 0.7189
Epoch 10/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8473 - accuracy: 0.7264
Epoch 11/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8195 - accuracy: 0.7317
Epoch 12/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8001 - accuracy: 0.7375
Epoch 13/100
47/47 [==============================] - 0s 956us/step - loss: 0.7797 - accuracy: 0.7378
Epoch 14/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7295 - accuracy: 0.7395
Epoch 15/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7197 - accuracy: 0.7546
Epoch 16/100
47/47 [==============================] - 0s 956us/step - loss: 0.7227 - accuracy: 0.7492
Epoch 17/100
47/47 [==============================] - 0s 934us/step - loss: 0.6953 - accuracy: 0.7550
Epoch 18/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6662 - accuracy: 0.7550
Epoch 19/100
47/47 [==============================] - 0s 2ms/step - loss: 0.6877 - accuracy: 0.7580
Epoch 20/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6857 - accuracy: 0.7476
Epoch 21/100
47/47 [==============================] - 0s 999us/step - loss: 0.6706 - accuracy: 0.7536
Epoch 22/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6530 - accuracy: 0.7583
Epoch 23/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6461 - accuracy: 0.7728
Epoch 24/100
47/47 [==============================] - 0s 999us/step - loss: 0.6464 - accuracy: 0.7597
Epoch 25/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6301 - accuracy: 0.7624
Epoch 26/100
47/47 [==============================] - 0s 2ms/step - loss: 0.6630 - accuracy: 0.7725
Epoch 27/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5932 - accuracy: 0.7785
Epoch 28/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6081 - accuracy: 0.7728
Epoch 29/100
47/47 [==============================] - 0s 999us/step - loss: 0.6017 - accuracy: 0.7738
Epoch 30/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6124 - accuracy: 0.7779
Epoch 31/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5681 - accuracy: 0.7937
Epoch 32/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6062 - accuracy: 0.7795
Epoch 33/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5817 - accuracy: 0.7816
Epoch 34/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5675 - accuracy: 0.7890
Epoch 35/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6193 - accuracy: 0.7839
Epoch 36/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5758 - accuracy: 0.7768
Epoch 37/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5754 - accuracy: 0.7802
Epoch 38/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5666 - accuracy: 0.7913
Epoch 39/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5717 - accuracy: 0.7906
Epoch 40/100
47/47 [==============================] - 0s 978us/step - loss: 0.5687 - accuracy: 0.7863
Epoch 41/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5666 - accuracy: 0.7893
Epoch 42/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5683 - accuracy: 0.7960
Epoch 43/100
47/47 [==============================] - 0s 913us/step - loss: 0.5641 - accuracy: 0.7933
Epoch 44/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5439 - accuracy: 0.8004
Epoch 45/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5920 - accuracy: 0.7863
Epoch 46/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5629 - accuracy: 0.7970
Epoch 47/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5574 - accuracy: 0.7980
Epoch 48/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5629 - accuracy: 0.7859
Epoch 49/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5459 - accuracy: 0.7994
Epoch 50/100
47/47 [==============================] - 0s 999us/step - loss: 0.5325 - accuracy: 0.8024
Epoch 51/100
47/47 [==============================] - 0s 913us/step - loss: 0.5386 - accuracy: 0.7991
Epoch 52/100
47/47 [==============================] - 0s 934us/step - loss: 0.5201 - accuracy: 0.8007
Epoch 53/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5178 - accuracy: 0.8034
Epoch 54/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5127 - accuracy: 0.8139
Epoch 55/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5472 - accuracy: 0.7933
Epoch 56/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5420 - accuracy: 0.7994
Epoch 57/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4740 - accuracy: 0.8159
Epoch 58/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5826 - accuracy: 0.8001
Epoch 59/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5207 - accuracy: 0.8044
Epoch 60/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5429 - accuracy: 0.8081
Epoch 61/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5102 - accuracy: 0.8102
Epoch 62/100
47/47 [==============================] - 0s 956us/step - loss: 0.5331 - accuracy: 0.8159
Epoch 63/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5056 - accuracy: 0.8152
Epoch 64/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5256 - accuracy: 0.8001
Epoch 65/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4929 - accuracy: 0.8095
Epoch 66/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5587 - accuracy: 0.7980
Epoch 67/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4815 - accuracy: 0.8172
Epoch 68/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5136 - accuracy: 0.8095
Epoch 69/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5189 - accuracy: 0.8145
Epoch 70/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5016 - accuracy: 0.8166
Epoch 71/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5018 - accuracy: 0.8176
Epoch 72/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5048 - accuracy: 0.8115
Epoch 73/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5167 - accuracy: 0.8115
Epoch 74/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5097 - accuracy: 0.8092
Epoch 75/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5022 - accuracy: 0.8152
Epoch 76/100
47/47 [==============================] - 0s 978us/step - loss: 0.4904 - accuracy: 0.8156
Epoch 77/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4918 - accuracy: 0.8125
Epoch 78/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4868 - accuracy: 0.8206
Epoch 79/100
47/47 [==============================] - 0s 999us/step - loss: 0.4981 - accuracy: 0.8176
Epoch 80/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5056 - accuracy: 0.8085
Epoch 81/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5025 - accuracy: 0.8108
Epoch 82/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5274 - accuracy: 0.8034
Epoch 83/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4852 - accuracy: 0.8166
Epoch 84/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4675 - accuracy: 0.8297
Epoch 85/100
47/47 [==============================] - 0s 978us/step - loss: 0.4943 - accuracy: 0.8156
Epoch 86/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4828 - accuracy: 0.8270
Epoch 87/100
47/47 [==============================] - 0s 999us/step - loss: 0.4784 - accuracy: 0.8169
Epoch 88/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4886 - accuracy: 0.8125
Epoch 89/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4946 - accuracy: 0.8162
Epoch 90/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4954 - accuracy: 0.8118
Epoch 91/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4670 - accuracy: 0.8219
Epoch 92/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5127 - accuracy: 0.8129
Epoch 93/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4565 - accuracy: 0.8206
Epoch 94/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4783 - accuracy: 0.8236
Epoch 95/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4732 - accuracy: 0.8156
Epoch 96/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4910 - accuracy: 0.8078
Epoch 97/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4649 - accuracy: 0.8256
Epoch 98/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4720 - accuracy: 0.8145
Epoch 99/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4894 - accuracy: 0.8216
Epoch 100/100
47/47 [==============================] - 0s 999us/step - loss: 0.4861 - accuracy: 0.8253
------------------------------------------------------------------------
Training for fold 4 ...
Epoch 1/100
47/47 [==============================] - 1s 1ms/step - loss: 7.6412 - accuracy: 0.5510
Epoch 2/100
47/47 [==============================] - 0s 999us/step - loss: 3.3996 - accuracy: 0.6109
Epoch 3/100
47/47 [==============================] - 0s 1ms/step - loss: 1.9571 - accuracy: 0.6452
Epoch 4/100
47/47 [==============================] - 0s 999us/step - loss: 1.4268 - accuracy: 0.6701
Epoch 5/100
47/47 [==============================] - 0s 1ms/step - loss: 1.2891 - accuracy: 0.6769
Epoch 6/100
47/47 [==============================] - 0s 999us/step - loss: 1.0999 - accuracy: 0.7028
Epoch 7/100
47/47 [==============================] - 0s 1ms/step - loss: 1.0694 - accuracy: 0.7011
Epoch 8/100
47/47 [==============================] - 0s 1ms/step - loss: 1.0431 - accuracy: 0.7018
Epoch 9/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9327 - accuracy: 0.7210
Epoch 10/100
47/47 [==============================] - 0s 2ms/step - loss: 0.9099 - accuracy: 0.7227
Epoch 11/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8939 - accuracy: 0.7301
Epoch 12/100
47/47 [==============================] - 0s 999us/step - loss: 0.8222 - accuracy: 0.7425
Epoch 13/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8229 - accuracy: 0.7459
Epoch 14/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8119 - accuracy: 0.7459
Epoch 15/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7969 - accuracy: 0.7402
Epoch 16/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7958 - accuracy: 0.7415
Epoch 17/100
47/47 [==============================] - 0s 956us/step - loss: 0.7861 - accuracy: 0.7536
Epoch 18/100
47/47 [==============================] - 0s 956us/step - loss: 0.7550 - accuracy: 0.7486
Epoch 19/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7156 - accuracy: 0.7600
Epoch 20/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7406 - accuracy: 0.7614
Epoch 21/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7234 - accuracy: 0.7553: 0s - loss: 0.7234 - accuracy: 0.75
Epoch 22/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6955 - accuracy: 0.7664
Epoch 23/100
47/47 [==============================] - 0s 934us/step - loss: 0.6972 - accuracy: 0.7627
Epoch 24/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6867 - accuracy: 0.7607
Epoch 25/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6878 - accuracy: 0.7617
Epoch 26/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7157 - accuracy: 0.7634
Epoch 27/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6141 - accuracy: 0.7802
Epoch 28/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6980 - accuracy: 0.7691
Epoch 29/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6341 - accuracy: 0.7657
Epoch 30/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6411 - accuracy: 0.7853
Epoch 31/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6674 - accuracy: 0.7718
Epoch 32/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7129 - accuracy: 0.7641
Epoch 33/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6139 - accuracy: 0.7802
Epoch 34/100
47/47 [==============================] - 0s 934us/step - loss: 0.6420 - accuracy: 0.7832
Epoch 35/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6637 - accuracy: 0.7728
Epoch 36/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6152 - accuracy: 0.7910
Epoch 37/100
47/47 [==============================] - 0s 999us/step - loss: 0.6614 - accuracy: 0.7718
Epoch 38/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6555 - accuracy: 0.7802
Epoch 39/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5914 - accuracy: 0.7812
Epoch 40/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6543 - accuracy: 0.7758
Epoch 41/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6284 - accuracy: 0.7832
Epoch 42/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6382 - accuracy: 0.7799
Epoch 43/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6135 - accuracy: 0.7822
Epoch 44/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6137 - accuracy: 0.7839
Epoch 45/100
47/47 [==============================] - 0s 934us/step - loss: 0.5912 - accuracy: 0.7930
Epoch 46/100
47/47 [==============================] - 0s 999us/step - loss: 0.6350 - accuracy: 0.7795
Epoch 47/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6426 - accuracy: 0.7789
Epoch 48/100
47/47 [==============================] - 0s 978us/step - loss: 0.5363 - accuracy: 0.8065
Epoch 49/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6796 - accuracy: 0.7742
Epoch 50/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6126 - accuracy: 0.7842
Epoch 51/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5876 - accuracy: 0.7906
Epoch 52/100
47/47 [==============================] - 0s 912us/step - loss: 0.6208 - accuracy: 0.7829
Epoch 53/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5663 - accuracy: 0.7943
Epoch 54/100
47/47 [==============================] - 0s 978us/step - loss: 0.5642 - accuracy: 0.8065
Epoch 55/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5562 - accuracy: 0.8034
Epoch 56/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5752 - accuracy: 0.8028
Epoch 57/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5654 - accuracy: 0.7974
Epoch 58/100
47/47 [==============================] - 0s 978us/step - loss: 0.6252 - accuracy: 0.7974
Epoch 59/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5547 - accuracy: 0.8001
Epoch 60/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6113 - accuracy: 0.7900
Epoch 61/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5885 - accuracy: 0.7933
Epoch 62/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5439 - accuracy: 0.8105
Epoch 63/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5594 - accuracy: 0.8085
Epoch 64/100
47/47 [==============================] - 0s 978us/step - loss: 0.5796 - accuracy: 0.8001
Epoch 65/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5992 - accuracy: 0.8004
Epoch 66/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5658 - accuracy: 0.7980
Epoch 67/100
47/47 [==============================] - 0s 999us/step - loss: 0.5938 - accuracy: 0.7947
Epoch 68/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5602 - accuracy: 0.8038
Epoch 69/100
47/47 [==============================] - 0s 999us/step - loss: 0.5690 - accuracy: 0.7984
Epoch 70/100
47/47 [==============================] - 0s 999us/step - loss: 0.5394 - accuracy: 0.8028
Epoch 71/100
47/47 [==============================] - 0s 934us/step - loss: 0.5976 - accuracy: 0.8024
Epoch 72/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5601 - accuracy: 0.8065
Epoch 73/100
47/47 [==============================] - 0s 999us/step - loss: 0.5598 - accuracy: 0.8034
Epoch 74/100
47/47 [==============================] - 0s 934us/step - loss: 0.5803 - accuracy: 0.8092
Epoch 75/100
47/47 [==============================] - 0s 999us/step - loss: 0.5403 - accuracy: 0.8189
Epoch 76/100
47/47 [==============================] - 0s 956us/step - loss: 0.5764 - accuracy: 0.8014
Epoch 77/100
47/47 [==============================] - 0s 978us/step - loss: 0.5279 - accuracy: 0.8122
Epoch 78/100
47/47 [==============================] - 0s 956us/step - loss: 0.5179 - accuracy: 0.8156
Epoch 79/100
47/47 [==============================] - 0s 999us/step - loss: 0.5650 - accuracy: 0.8034
Epoch 80/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5283 - accuracy: 0.8149
Epoch 81/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5628 - accuracy: 0.8048
Epoch 82/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5347 - accuracy: 0.8149
Epoch 83/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5798 - accuracy: 0.8007
Epoch 84/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5245 - accuracy: 0.8162
Epoch 85/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5400 - accuracy: 0.8145
Epoch 86/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5434 - accuracy: 0.8139
Epoch 87/100
47/47 [==============================] - 0s 934us/step - loss: 0.5230 - accuracy: 0.8145
Epoch 88/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5322 - accuracy: 0.8152
Epoch 89/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5290 - accuracy: 0.8216
Epoch 90/100
47/47 [==============================] - 0s 999us/step - loss: 0.5526 - accuracy: 0.8108
Epoch 91/100
47/47 [==============================] - 0s 934us/step - loss: 0.5760 - accuracy: 0.8092
Epoch 92/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5270 - accuracy: 0.8135
Epoch 93/100
47/47 [==============================] - 0s 913us/step - loss: 0.5271 - accuracy: 0.8199
Epoch 94/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5064 - accuracy: 0.8182
Epoch 95/100
47/47 [==============================] - 0s 978us/step - loss: 0.5598 - accuracy: 0.8095
Epoch 96/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5615 - accuracy: 0.8081
Epoch 97/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5005 - accuracy: 0.8139
Epoch 98/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5045 - accuracy: 0.8236
Epoch 99/100
47/47 [==============================] - 0s 934us/step - loss: 0.4957 - accuracy: 0.8294
Epoch 100/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5254 - accuracy: 0.8179
------------------------------------------------------------------------
Training for fold 5 ...
Epoch 1/100
47/47 [==============================] - 1s 978us/step - loss: 14.9073 - accuracy: 0.4095
Epoch 2/100
47/47 [==============================] - 0s 1ms/step - loss: 3.8147 - accuracy: 0.4990
Epoch 3/100
47/47 [==============================] - 0s 1ms/step - loss: 2.0952 - accuracy: 0.5511
Epoch 4/100
47/47 [==============================] - 0s 1ms/step - loss: 1.4645 - accuracy: 0.6036
Epoch 5/100
47/47 [==============================] - 0s 1ms/step - loss: 1.2158 - accuracy: 0.6423
Epoch 6/100
47/47 [==============================] - 0s 1ms/step - loss: 1.0493 - accuracy: 0.6669
Epoch 7/100
47/47 [==============================] - 0s 1ms/step - loss: 0.9792 - accuracy: 0.6817
Epoch 8/100
47/47 [==============================] - 0s 934us/step - loss: 0.8948 - accuracy: 0.6962
Epoch 9/100
47/47 [==============================] - 0s 1ms/step - loss: 0.8687 - accuracy: 0.7046
Epoch 10/100
47/47 [==============================] - 0s 956us/step - loss: 0.8181 - accuracy: 0.7201
Epoch 11/100
47/47 [==============================] - 0s 999us/step - loss: 0.7998 - accuracy: 0.7295
Epoch 12/100
47/47 [==============================] - 0s 956us/step - loss: 0.8257 - accuracy: 0.7184
Epoch 13/100
47/47 [==============================] - 0s 999us/step - loss: 0.7834 - accuracy: 0.7332
Epoch 14/100
47/47 [==============================] - 0s 934us/step - loss: 0.7753 - accuracy: 0.7325
Epoch 15/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7064 - accuracy: 0.7470
Epoch 16/100
47/47 [==============================] - 0s 934us/step - loss: 0.8075 - accuracy: 0.7318
Epoch 17/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7270 - accuracy: 0.7483
Epoch 18/100
47/47 [==============================] - 0s 934us/step - loss: 0.7297 - accuracy: 0.7527
Epoch 19/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6969 - accuracy: 0.7453
Epoch 20/100
47/47 [==============================] - 0s 999us/step - loss: 0.7207 - accuracy: 0.7513
Epoch 21/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6533 - accuracy: 0.7651
Epoch 22/100
47/47 [==============================] - 0s 913us/step - loss: 0.7033 - accuracy: 0.7645
Epoch 23/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6993 - accuracy: 0.7571
Epoch 24/100
47/47 [==============================] - 0s 1ms/step - loss: 0.7054 - accuracy: 0.7608
Epoch 25/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6307 - accuracy: 0.7695
Epoch 26/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6609 - accuracy: 0.7688
Epoch 27/100
47/47 [==============================] - 0s 978us/step - loss: 0.6256 - accuracy: 0.7746
Epoch 28/100
47/47 [==============================] - 0s 978us/step - loss: 0.6557 - accuracy: 0.7719
Epoch 29/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6510 - accuracy: 0.7769
Epoch 30/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6841 - accuracy: 0.7688
Epoch 31/100
47/47 [==============================] - 0s 913us/step - loss: 0.6115 - accuracy: 0.7742
Epoch 32/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6445 - accuracy: 0.7705
Epoch 33/100
47/47 [==============================] - 0s 956us/step - loss: 0.5861 - accuracy: 0.7900
Epoch 34/100
47/47 [==============================] - ETA: 0s - loss: 0.5856 - accuracy: 0.81 - 0s 1ms/step - loss: 0.6471 - accuracy: 0.7789
Epoch 35/100
47/47 [==============================] - 0s 978us/step - loss: 0.6367 - accuracy: 0.7732
Epoch 36/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6419 - accuracy: 0.7766
Epoch 37/100
47/47 [==============================] - 0s 934us/step - loss: 0.5980 - accuracy: 0.7921
Epoch 38/100
47/47 [==============================] - 0s 999us/step - loss: 0.6136 - accuracy: 0.7867
Epoch 39/100
47/47 [==============================] - 0s 978us/step - loss: 0.5909 - accuracy: 0.7857
Epoch 40/100
47/47 [==============================] - 0s 913us/step - loss: 0.5844 - accuracy: 0.7988
Epoch 41/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6201 - accuracy: 0.7900
Epoch 42/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5940 - accuracy: 0.7803
Epoch 43/100
47/47 [==============================] - 0s 1ms/step - loss: 0.6070 - accuracy: 0.7884
Epoch 44/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5597 - accuracy: 0.8008
Epoch 45/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5993 - accuracy: 0.7904
Epoch 46/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5479 - accuracy: 0.8059
Epoch 47/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5335 - accuracy: 0.7985
Epoch 48/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5805 - accuracy: 0.7988
Epoch 49/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5519 - accuracy: 0.7981
Epoch 50/100
47/47 [==============================] - 0s 2ms/step - loss: 0.6055 - accuracy: 0.7998
Epoch 51/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5494 - accuracy: 0.7978
Epoch 52/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5600 - accuracy: 0.7974
Epoch 53/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5329 - accuracy: 0.8122
Epoch 54/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5465 - accuracy: 0.8038
Epoch 55/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5619 - accuracy: 0.8045
Epoch 56/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5365 - accuracy: 0.8102
Epoch 57/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5848 - accuracy: 0.7964
Epoch 58/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5489 - accuracy: 0.8092
Epoch 59/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5119 - accuracy: 0.8092
Epoch 60/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5492 - accuracy: 0.8022
Epoch 61/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5258 - accuracy: 0.8082
Epoch 62/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5350 - accuracy: 0.8146
Epoch 63/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5460 - accuracy: 0.8085
Epoch 64/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5123 - accuracy: 0.8173
Epoch 65/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5300 - accuracy: 0.8119
Epoch 66/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5278 - accuracy: 0.8102
Epoch 67/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5291 - accuracy: 0.8116
Epoch 68/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5385 - accuracy: 0.8210
Epoch 69/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4963 - accuracy: 0.8230
Epoch 70/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5398 - accuracy: 0.8052
Epoch 71/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5465 - accuracy: 0.8116
Epoch 72/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5415 - accuracy: 0.8082
Epoch 73/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5221 - accuracy: 0.8136
Epoch 74/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5398 - accuracy: 0.8079
Epoch 75/100
47/47 [==============================] - 0s 2ms/step - loss: 0.4747 - accuracy: 0.8274
Epoch 76/100
47/47 [==============================] - 0s 2ms/step - loss: 0.5321 - accuracy: 0.8079
Epoch 77/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4785 - accuracy: 0.8324
Epoch 78/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5028 - accuracy: 0.8136
Epoch 79/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5272 - accuracy: 0.8166
Epoch 80/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4766 - accuracy: 0.8321
Epoch 81/100
47/47 [==============================] - 0s 999us/step - loss: 0.5462 - accuracy: 0.8136
Epoch 82/100
47/47 [==============================] - 0s 891us/step - loss: 0.4726 - accuracy: 0.8271
Epoch 83/100
47/47 [==============================] - 0s 934us/step - loss: 0.5114 - accuracy: 0.8183
Epoch 84/100
47/47 [==============================] - 0s 978us/step - loss: 0.4994 - accuracy: 0.8254
Epoch 85/100
47/47 [==============================] - 0s 978us/step - loss: 0.4902 - accuracy: 0.8311
Epoch 86/100
47/47 [==============================] - 0s 999us/step - loss: 0.5145 - accuracy: 0.8200
Epoch 87/100
47/47 [==============================] - 0s 978us/step - loss: 0.4885 - accuracy: 0.8294
Epoch 88/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4887 - accuracy: 0.8220
Epoch 89/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5066 - accuracy: 0.8294
Epoch 90/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4869 - accuracy: 0.8240
Epoch 91/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4511 - accuracy: 0.8331
Epoch 92/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5053 - accuracy: 0.8200
Epoch 93/100
47/47 [==============================] - 0s 1ms/step - loss: 0.5083 - accuracy: 0.8297
Epoch 94/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4995 - accuracy: 0.8193
Epoch 95/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4999 - accuracy: 0.8247
Epoch 96/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4994 - accuracy: 0.8163
Epoch 97/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4966 - accuracy: 0.8234
Epoch 98/100
47/47 [==============================] - 0s 956us/step - loss: 0.4639 - accuracy: 0.8297
Epoch 99/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4956 - accuracy: 0.8260
Epoch 100/100
47/47 [==============================] - 0s 1ms/step - loss: 0.4795 - accuracy: 0.8318
------------------------------------------------------------------------
Training for fold 6 ...
Epoch 1/100