Skip to content

Latest commit

 

History

History
56 lines (52 loc) · 1.5 KB

README.md

File metadata and controls

56 lines (52 loc) · 1.5 KB
Raw

Recommender System Library

Recommender System Library is a Python implementation of a number of popular recommendation algorithms for both implicit and explicit feedback.

Install

First, you need install Numpy, Pandas, Scikit Learn, scipy Sparse, Surpise, Implicit, LightFM

Then, clone repository and install using pip

$ git clone https://github.com/celerative/PPS-RecSys-Nicolas.git
$ pip install -e .

Features

Models

Recomendation Model

  • Alternating Least Squares (ALS)

Prediction Model

  • Bayesian Personalized Ranking (BPR)
  • Global Average
  • Item Average
  • User Average
  • K-Nearest Neighbors (KNN)
    • KNN Basic
    • KNN Baseline
    • KNN with Means
    • KNN with Zero Score
  • Single Value Descomposition (SVD)
  • Weighted Approximate-Rank Pairwise (WARP)

Accuracy

For Regression Model

  • Mean Absolute Error (MAE)
  • Root Mean Square Error (RMSE)
  • Fraction of Concordant Pairs (FCP)

For Classification Model

  • Area Under the Curve (AUC)
  • Precision
  • Recall
  • F1

Metrics

  • Cross Validation
  • Grid Search CV

Model Selection

  • Cosine Similarity
  • Pearson Correlation

Persistence

  • Load: load model
  • Dump: save model

Support

  • Marcos, Federico
  • Palazzesi, Nicolás