Skip to content

This is a machine learning course project led by me and developed by me and other four team members. We used R as it was stipulated by the course otherwise we would use Python. The model we built was an ensemble of XGB --- by training ten XGB models each of which learnt a distinct fraction of the dataset.

Notifications You must be signed in to change notification settings

Chacoon3/Airbnb_predictive_analysis

Repository files navigation

Airbnb_predictive_analysis

For Developers and Researchers

Please go check the file ./Library/utils.R to see a set of utility functions that I (Chacoon3) wrote to speed up the machine learning workflow.

Introduction

  • This is the winner machine learning project of my master program's machine learning contest. It contains visualizations, datasets, documents, team-developed R libraries, trained models, as well as source codes.

  • In this project, the repository owner along with other four teammates collaborated to train machine learning models that best ranks the popularity of Airbnb accomodations as were measured by a categorical target variable high_booking_rate.

  • The most important codes of this project are in three files, namely main.r, utils.r, and data_cleaning.r. The latter two files are under the folder "Library".

  • We finalized with a XGBoost model which had its AUC being 0.903, the highest AUC among all the teams that parcipated in this competition.

About

This is a machine learning course project led by me and developed by me and other four team members. We used R as it was stipulated by the course otherwise we would use Python. The model we built was an ensemble of XGB --- by training ten XGB models each of which learnt a distinct fraction of the dataset.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages