email: [email protected]
Welcome to this project! It's part of Henry Bootcamp Labs Individual Projects.
You can check the repository and the readme on this link: https://github.com/soyHenry/Datathon or at CONSIGNAS.md file.
For further analysis and development forwarding this contest everything is written in English.
As you can see this repository has 5 main notebooks. You can open and get into the dataset transformation and analysis running each one in order.
- Notebook 1: Initial cleaning, feature extraction, imputation and analysis.
- Notebook 2: Same approach for test set
- Notebook 3: GPS location feature engineering
- Notebook 4: Machine learning models with special focus on ensemble tree models
- Notebook 5: Different approach with a recommendation system with CountVectorizer
googlemapsapi: You must open it first and extract the missing GPS coordinates from Google Maps API. Data is already downloaded in /data path but in case you wanna try you should first get an ACCESS_KEY for the Google API.
Hope you enjoy!