https://github.com/julscadenas/ds-ml-projects

Collection of my Data Science and Machine Learning Projects.
https://github.com/julscadenas/ds-ml-projects

data-science machine-learning python

Last synced: about 2 months ago
JSON representation

Collection of my Data Science and Machine Learning Projects.

• Host: GitHub
• URL: https://github.com/julscadenas/ds-ml-projects
• Owner: julsCadenas
• Created: 2025-02-10T14:45:58.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-24T14:11:06.000Z (over 1 year ago)
• Last Synced: 2025-02-24T15:27:30.684Z (over 1 year ago)
• Topics: data-science, machine-learning, python
• Language: Jupyter Notebook
• Homepage:
• Size: 8.62 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# **Data Science & Machine Learning Project-Based Learning Plan**

This repository contains my hands-on projects as part of my Machine Learning Project-Based Learning Plan. Each project follows a structured learning approach, covering various ML concepts and techniques.

## 📖 Table of Contents  

1. [🏡 House Price Prediction (Regression)](https://github.com/julsCadenas/DS-ML-Projects/tree/main/HousePricePrediction)  

2. [📧 Spam Email Classifier](https://github.com/julsCadenas/DS-ML-Projects/tree/main/SpamEmailClassifier)  

3. [📺 Investigating Netflix Movies (DataCamp)](https://github.com/julsCadenas/DS-ML-Projects/tree/main/InvestigateNetflixMovies)

4. [🙍‍♂️ Customer Churn Prediction](https://github.com/julsCadenas/DS-ML-Projects/tree/main/CustomerChurnPrediction)

5. [👮 Investigating Crime in Los Angeles (DataCamp)](https://github.com/julsCadenas/DS-ML-Projects/tree/main/CrimesInLA)

6. [🌾 Predictive Modeling for Agriculture (DataCamp)](https://github.com/julsCadenas/DS-ML-Projects/tree/main/PredictiveModelAgriculture)

7. [🌡️ Predicting Temperature in London (DataCamp)](https://github.com/julsCadenas/DS-ML-Projects/tree/main/LondonTemps)

8. [🐤 Sentiment Analysis on Tweets](https://github.com/julsCadenas/DS-ML-Projects/tree/main/TwtSentimentAnalysis)

9. [1️⃣ MNIST Handwritten Digit Classifier](https://github.com/julsCadenas/DS-ML-Projects/tree/main/TFMnist)

---

### **🏡 House Price Prediction (Regression)**  

- **Dataset:** [Kaggle: House Prices - Advanced Regression Techniques](https://www.kaggle.com/competitions/house-prices-advanced-regression-techniques)  

- **Key Concepts:** Regression, Feature Engineering, Hyperparameter Tuning  

- **Tools Used:** Scikit-Learn, Pandas, Matplotlib, NumPy  

- **Summary:**  

  - Used **Linear Regression & Lasso Regression** to predict house prices.  

  - Performed **EDA**, **feature engineering**, and **hyperparameter tuning**.  

  - Evaluated model using **RMSE, MAE, and R² score**.  

- **Status:** ✅ Completed  

- **Code & Notebook:** [House Price Prediction](https://github.com/julsCadenas/DS-ML-Projects/tree/main/HousePricePrediction)  

---

### **📧 Spam Email Classifier (Classification)**  

- **Dataset:** [Kaggle: Email Spam Detection Dataset](https://www.kaggle.com/datasets/shantanudhakadd/email-spam-detection-dataset-classification)  

- **Key Concepts:** Naïve Bayes, NLP, Text Preprocessing  

- **Tools Used:** Scikit-Learn, NLTK, Pandas, Matplotlib, Seaborn, NumPy  

- **Summary:**  

  - Trained and compared **Multinomial Naïve Bayes**, **Bernoulli Naïve Bayes**, and **MLP Classifier**.  

  - Evaluated performance using **Accuracy, Precision, Recall, and F1-score**.  

- **Status:** ✅ Completed  

- **Code & Notebook:** [Spam Email Classifier](https://github.com/julsCadenas/DS-ML-Projects/tree/main/SpamEmailClassifier)  

---

### **📺 Investigating Netflix Movies (DataCamp)**

- **Dataset:** [Investigating Netflix Movies](https://app.datacamp.com/learn/projects/investigating_netflix/guided/Python)  

- **Key Concepts:** Exploratory Data Analysis  

- **Tools Used:** Pandas, Matplotlib  

- **Summary:**  

  - Perform Exploratory Data Analysis on the dataset

- **Status:** ✅ Completed 

- **Code & Notebook:** [Netflix Data EDA](https://github.com/julsCadenas/DS-ML-Projects/tree/main/InvestigateNetflixMovies)  

---

### **🙍‍♂️ Customer Churn Prediction**

- **Datasets:**

  - [Customer Churn Dataset](https://www.kaggle.com/datasets/muhammadshahidazeem/customer-churn-dataset)  

  - [Telco Churn Dataset](https://www.kaggle.com/datasets/mnassrib/telecom-churn-datasets)  

  - [Bank Customer Churn Dataset](https://www.kaggle.com/datasets/gauravtopre/bank-customer-churn-dataset)  

- **Key Concepts:** Classification, Model Evaluation, EDA  

- **Tools Used:** XGBoost, Scikit-Learn, SHAP, Pandas, Matplotlib, Seaborn  

- **Summary:**  

  - Trained XGBoost model for churn prediction.

  - Evaluated using Accuracy, Precision, Recall, F1-score, and AUC-ROC.

  - Used SHAP analysis to interpret feature importance (e.g., payment delays, support calls).

  - Achieved 91% recall with an AUC-ROC of 0.9006.

- **Status:** ✅ Completed

- **Code & Notebook:** [Customer Churn Pred](https://github.com/julsCadenas/DS-ML-Projects/tree/main/CustomerChurnPrediction)  

---

### **👮 Analyzing Crime in Los Angeles (DataCamp)**

- **Dataset:** [Analyzing Crime in LA](https://app.datacamp.com/learn/projects/investigating_netflix/guided/Python)

- **Key Concepts:** Exploratory Data Analysis

- **Tools used:** Pandas, Matplotlib, Seaborn

- **Summary:**

  - Perform EDA on the dataset

- **Status:** ✅ Completed

- **Code & Notebook:** [Crime in LA](https://github.com/julsCadenas/DS-ML-Projects/tree/main/CrimesInLA)

---

### **🌾 Predictive Modeling for Agriculture (DataCamp)**

- **Dataset:** [Predictive Model Agriculture](https://app.datacamp.com/learn/projects/1772)

- **Key Concepts:** Exploratory Data Analysis, Logistic Regression, Model Evaluation

- **Tools used:** Pandas, Scikit-Learn

- **Summary:**

  - Perform EDA on the dataset

  - Use Logistic Regression on each feature.

  - Find the most performant feature.

  - Evaluate the model using F1 score.

- **Status:** ✅ Completed

- **Code & Notebook:** [Agriculture Model](https://github.com/julsCadenas/DS-ML-Projects/tree/main/PredictiveModelAgriculture)

---

### **🌡️ Predicting Temperature in London (DataCamp)**

- **Dataset:** [Predicting Temperature in London](https://app.datacamp.com/learn/projects/predicting_temperature_in_london/guided/Python)

- **Key Concepts:** Exploratory Data Analysis, Linear Regression, Decision Trees, Random Forest, Model Evaluation, MLflow

- **Tools used:** Pandas, Numpy, Scikit-Learn, MLflow

- **Summary:**

  - Load and clean historical weather data from London.

  - Perform EDA to understand seasonal trends and feature correlations.

  - Select features relevant for predicting mean temperature.

  - Preprocess the data using imputation and scaling.

  - Train and evaluate Linear Regression, Decision Tree, and Random Forest models.

  - Log models, parameters, and performance metrics using MLflow.

- **Status:** ✅ Completed

- **Code & Notebook:** [London Temps](https://github.com/julsCadenas/DS-ML-Projects/tree/main/LondonTemps)

---

### **🐤 Sentiment Analysis on Tweets**

- **Dataset:** [Sentiment140 Dataset](https://www.kaggle.com/datasets/kazanova/sentiment140/code)

- **Key Concepts:** Natural Language Processing (NLP), Text Processing, Binary Classification, Model Evaluation, Long Short-Term Memory (LSTM) 

- **Tools used:**  Pandas, Numpy, Scikit-Learn, NLTK, Matplotlib, Seaborn, Tensorflow, Keras

- **Summary:**

  - Loaded and cleaned 1.6 million tweets labeled as positive or negative.

  - Applied text preprocessing: lowercasing, punctuation and stopword removal, tokenization.

  - Converted text to sequences using Keras `Tokenizer`, and padded them to equal length.

  - Built a neural network using TensorFlow/Keras with embedding and dense layers.

  - Trained the model and evaluated it on a 320,000-sample test set.

  - Achieved the following evaluation metrics:

    - **Accuracy:** 81.12%

    - **Precision:** 81.26%

    - **Recall:** 80.96%

    - **F1-Score:** 81.11%

- **Status:** ✅ Completed

- **Code & Notebook:** [Tweet Sentiment Analysis](https://github.com/julsCadenas/DS-ML-Projects/tree/main/TwtSentimentAnalysis)

---

### **1️⃣ MNIST Handwritten Digit Recognition**

- **Dataset:** MNIST (from `tf.keras.datasets`)

- **Key concepts:** Neural networks, image classification, data normalization, softmax & reLU activation, model evaluation

- **Tools used:** TensorFlow, Keras, matplotlib, numpy

- **Summary:**

  - Built a fully connected neural network (FCNN) to classify 28x28 grayscale images of handwritten digits.

  - Preprocessed the dataset by normalizing pixel values; trained over 5 epochs and visualized predictions.

  - Evaluated the model using accuracy and loss metrics on 10,000 test images.

- **Status:** ✅ Completed

- **Code & Notebook:** [TFMnist](https://github.com/julsCadenas/DS-ML-Projects/tree/main/TFMnist)