Game Recommender

Steam Game Recommender

Overview

The project includes Python scripts for various tasks such as importing data, preprocessing, partitioning data, computing similarities, training models, and making predictions. It consists of two main parts:

1. Game Prediction:

   • Predict whether a user has played a particular game.
   • Build models, including logistic regression, for game prediction.
   • Evaluate model performance and test on provided data.

2. Hours Played Prediction:

   • Predict the number of hours a user has played a game.
   • Perform preprocessing steps and define functions for iteration.
   • Train the model and make predictions on test data.

File Structure

Data/
    train.json.gz
    pairs_Played.csv
    pairs_Hours.csv
README.md
assignment1.py
predictions_Played.csv
predictions_Hours.csv

Usage

1. Data Preparation:

   • Place the provided data files (train.json.gz, pairs_Played.csv, pairs_Hours.csv) in the Data/ directory.

2. Running the Code:

   • Execute the assignment1.py script to run the entire analysis pipeline.
   • Ensure all required libraries are installed (gzip, scipy, sklearn, numpy, etc.).

3. Viewing Results:

   • The predictions for game plays (predictions_Played.csv) and hours played (predictions_Hours.csv) will be generated.
   • Explore the results and evaluate model performance based on accuracy metrics.