This project involves the development of a Python-based Machine Learning-based Recommender System focused on predicting user preferences using video game reviews sourced from Steam:
- The initial segment encompasses a play prediction recommender utilizing Bayesian Personalized Ranking (BPR) and TensorFlow to forecast a user's likelihood of engaging with a specific game or item based on their historical play data.
- Subsequently, a time prediction recommender is implemented, employing a variation of Stochastic Gradient Descent (SGD) to anticipate the duration a user might spend playing a particular game or item, relying on their historical playtime data.