santiagoahl / platzi-world-happiness

A Machine Learning Model built in scikit-learn using Support Vector Regressors, Ensemble modeling with Gradient Boost Regressor and Grid Search Cross Validation.

World Happiness Report

A Machine Learning Model API in scikit-learn using Support Vector Regressors and ensemble modeling with Gradient Boost Regressor and Cross Validation.

Key Features • How To Use • Credits • License

Key Features

This machine learning model predicts the happines score of a given country. This prediction is a number between 0 and 10. The dataset is taken from the World Happiness Report Kaggle Competition. So here are the key features of this project:

• Prediction is based on this country's features:
  • high
  • low
  • gdp
  • family
  • lifexp
  • freedom
  • generosity
  • corruption
  • dystopia : Imaginary country that has the world's least-happy people.
• Professional Modularization on this Project. Some modules are programmed using OOP.
• Built with an Rest API programmed in Flask .
• Based on Scikit-Learn modules and functions such like:
  • svm.SVR : Support Vector Regressor.
  • ensemble.GradientBoostingRegressor : Gradiente Boosting Regressors Ensemble method.
  • model_selection.GridSearchCV : Cross validation method.

How To Use

To clone and run this application, follow these steps

# Clone this repository
$ git clone https://github.com/santiagoahl/world-happiness.git

# Go into the repository
$ cd world-happiness

# Install dependencies
$ pip install -r requirements.txt

# Run the app
$ python3 server.py

#View results putting the following on your browser (If port 8080 is busy change it)

http://127.0.0.1:8080/predict

Credits

This software uses the following packages:

• Scikit-Learn
• Flask
• Joblib

License

MIT

---

Web Site santiagoal.super.site  ·  GitHub @santiagoahl  ·  Twitter @sahumadaloz