Polimi Recommender System challenge A.Y. 2022/23

This repo contains the code and the data used in polimi's recsys challenge

Results

Deadline 2 (final)

• Public leaderboard: 2th
• Private leaderboard: 3th

Deadline 1

• Public leaderboard: 5th
• Private leaderboard: 1st

MAP@10: 0.06200

Best model

The final model is a hybrid recommender composed of

1. A hybrid recommender trained on a binary URM
2. A hybrid recommender trained on a real-valued URM
3. IALS on binary URM from Implicit's library

The first hybrid is composed of

• SLIM-ElasticNet (stacked URM)
• IALS
• MultVAE
• Mini hybrid of RP3Beta (user-similarity + item-similarity)
• EASE-R (weighted stacked URM)

The second hybrid is composed of

• EASE-R
• SLIM-ElasticNet
• RP3Beta (stacked)
• IALS_implicit
• MultVAE

This repository

This repository is based on Maurizio Dacrema's repository

About us

Andrea Riboni and Nicola Cecere

My telegram

Tuning

• Macbook Pro 2019
• Lenovo Yoga Slim 7 Pro
• Standard E2s v3 (2 vcpus, 16 GiB memory) Azure VM, Ubuntu