LogoAlan Craig Data Science Portfolio Architecture

/*

Recommender Systems: Collaborative Filtering

python logo

In this project, I create a proof of concept web application that showcases different movie recommendation algorithms. In this project, you will build a proof of concept: A web application that showcases different movie recommendation algorithms.

The first 3 steps involve:

Collaborative filtering (CF) is an umbrella term for various methods related to recommender systems. Despite their differences, all methods in CF use implicit or explicit ratings as input for creating recommendations. The central assumption is that if two users have rated some items similarly or showed similar behavior in the past, they will also act similar on future, possibly unseen, items.

Non-negative matrix factorization is another unsupervised learning algorithm that aims to extract useful features, by decomposing data to find latent factors which can be described by matrices. Although it is classified as an unsupervised learning technique, this statistical model can be used to learn and predict ratings.

Pick and implement a Collaborative Filtering recommender using Collaborative Filtering with Matrix. These steps involve:

Contact

Lets get in touch and talk about your next project.

>