yifanjiang19 / purchaseprediction

2017 Seedcup Semi-final Competition

Purchase Prediction

2017 Seed Cup Semi-Final Competition

Introduction

1. Firstly, we need to extract inherent features and of users and products. Furtherly, we extracted behavioural features of users and the interactive features of products all from user's behaviors

2. We developed an User-Product Purchasing Model

3. We adopt user-based and product-based collaborative filtering to generate recommend products for users

Environment

• macOS 10.14 High Sierra

• Python 2.7.10

• XGBoost

• Pytorch

• Scikit-Learn

Statement

The code commited on this repository might not work much well as expected, because we adopted many intermediate steps and operations to collect the best result!

Licence

This project is under the MIT Licence