https://github.com/recombee/sparseelsa

https://github.com/recombee/sparseelsa

Last synced: 2 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/recombee/sparseelsa
• Owner: recombee
• License: other
• Created: 2024-02-08T21:06:39.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2025-11-12T07:16:18.000Z (7 months ago)
• Last Synced: 2025-11-12T09:16:22.543Z (7 months ago)
• Language: Python
• Size: 52.5 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# SparseELSA 

[![License: CC BY-NC 4.0](https://img.shields.io/badge/License-CC_BY--NC_4.0-lightgrey.svg)](https://creativecommons.org/licenses/by-nc/4.0/)

Forked from [Recommender System Datasets (and Baselines)](https://github.com/zombak79/recsysdata)

1. Download datasets. For example, for Movielens20m dataset go to datasets/ml20m folder and run `download_ml20m` script.

2. Install dependencies `pip install -r requirements.txt`

3. Run desired experiment. For example for Sparse ELSA run `python experiment_elsa.py --dataset ml20m --validation false --factors 800 --batch_size 1024 --model_strategy super_sparse --max_output 20000 --scheduler none --epochs 10 --tuning False --flag elsa_ml20m_test`

4. Run `python analyze_results.py -f elsa_ml20m_test` to see results of the experiment:

```

  dataset             flag  recall@20  recall@50  ndcg@100  training_time  std_recall@20  std_recall@50  std_ndcg@100  std_time  n_experiments

0   ml20m  elsa_ml20m_test   0.391659   0.528692  0.430469      32.278537            NaN            NaN           NaN       NaN              1

```