https://github.com/khuyentran1401/suicide-rates

https://github.com/khuyentran1401/suicide-rates

data-analysis data-science kaggle machine-learning python

Last synced: 1 day ago
JSON representation

• Host: GitHub
• URL: https://github.com/khuyentran1401/suicide-rates
• Owner: khuyentran1401
• Created: 2019-11-26T13:52:24.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2020-04-06T15:25:42.000Z (almost 5 years ago)
• Last Synced: 2024-11-26T11:50:07.668Z (2 months ago)
• Topics: data-analysis, data-science, kaggle, machine-learning, python
• Language: Jupyter Notebook
• Size: 805 KB
• Stars: 3
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# About this Project

Find the predictors of suicide among different countries, years, sex, generation, and age group.

# Data

The data is retrieved from [Kaggle Suicide Rates Overview 1985 to 2016](https://www.kaggle.com/russellyates88/suicide-rates-overview-1985-to-2016)

# Notebook

Access the notebook of this project [here](https://github.com/khuyentran1401/Suicide-rates/blob/master/suicide-rates-1985-2016.ipynb)

# Tools

* Numpy, Pandas

* Seaborn

* Scikit-learn

# Steps

1. Import and explore data

1. Preprocess data

    * Split traing and test set with `sklearn.model_selection.train_test_split`

    * Shuffle the data to increase randomness with `sklearn.model_selection.StratifiedShuffleSplit`

    * Merge map data to existing data

    * Drop the unimportant features

    * Transform categorical data with `sklearn.preprocessing.LabelEncoder`

    * Impute missing data with `sklearn.preprocessing.Imputer`

    * Scale data with `sklearn.preprocessing.StandardScaler`

1. Visualization

    * Countplot to find the distribution of data

    * Heatmap and pairplot to find the correlation between features

    * Geopandas to find the distribution of suicide rates accross countries in the world

     

    ![image](https://github.com/khuyentran1401/Suicide-rates/blob/master/images/Screenshot%202020-04-06%2010.20.56.png)

1. Metrics:

    * Mean squared error

    * Cross validation score

1. Model Training

    * Linear Regression

    * Decision Tree Regressor

    * Random Forest Regressor

    * Grid Search Cross Validation

    * Randomized Search CV

    

    ![image](https://github.com/khuyentran1401/Suicide-rates/blob/master/images/Screenshot%202020-04-06%2010.13.21.png?raw=true)

    

    _Compare hyperparameters for Randomized Search CV_