Collaborative Filtering – Movies Recommendation

This group project aims to predict missing entries in a movie-user ratings matrix from the MovieLens dataset. The codebase is organized into modular components for data preprocessing, matrix completion algorithms, evaluation metrics, and utilities.

Documents

For context and overview of the work achieved one can refer to these resources :

• Instructions slides (instructions.pdf)
• Short report presenting our main ideas and results (report.pdf)
• Slides for the project presentation (slides.pdf)

Requirements

• Python 3.8+
• Dependencies: run pip install -r requirements.txt from your desired environment

Usage

1. Place your MovieLens sparse users/movies rating matrix in the data/ directory (or any other directory)
2. Run the main script:

python generate.py --name "data/sparse_matrix"

Core Modules

metrics

Evaluation metrics for matrix completion performance.

Functions:

• rmse(pred_matrix, true_matrix): Root Mean Square Error calculation
• accuracy_exact(pred_matrix, true_matrix): Exact accuracy (rounded predictions)

preprocessing

Data preprocessing utilities for matrix completion tasks.

DataPreprocessor class methods:

• fusion(): Combine training and test matrices
• split(): Split ratings into train/test sets
• normalize_by_user(): User-centered normalization
• denormalize_by_user(): Restore original scale
• filter_by_threshold(): Remove users/movies with few ratings

matrix_completion_methods

Matrix completion algorithms implementing different matric completion approaches :

• AverageCompletion: Simple baseline method using row/column averages to fill missing values.

• MatrixFactorisation: Advanced method using matrix factorization with two different algorithms (Alternating Least Squares and Gradient-based optimization)

• IterativePCA: Iterative PCA-based imputation that alternates between estimating missing entries and computing a low-rank PCA reconstruction until convergence.

• IterativeKernelPCA: Extension of the previous method, using kernel techniques to try to capture nonlinear behavior in the data.

tuning

Utilities for hyper-parameter selection and model validation. Provides basic cross validation and tuning tools such as K-folds and grid search.