Skip to content

EDA and machine learning for a Colombia real state dataset. Using geospatial data for feature engineering.

Notifications You must be signed in to change notification settings

maicobernal/real_state_colombia

Repository files navigation

Datathon Henry Labs 2 - Cohorte 04

03/11/2022

Autor: Maico Bernal

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!

About

EDA and machine learning for a Colombia real state dataset. Using geospatial data for feature engineering.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published