From the course: Building Recommender Systems with Machine Learning and AI

Unlock this course with a free trial

Join today to access over 22,600 courses taught by industry experts.

Tuning collaborative filtering algorithms

Tuning collaborative filtering algorithms - Python Tutorial

From the course: Building Recommender Systems with Machine Learning and AI

Start my 1-month free trial Buy for my team

Tuning collaborative filtering algorithms

“

- [Instructor] Let's fiddle with our results a bit because as we've said, there are many ways to implement user-based and item-based collaborative filtering. One thing we're doing that's kind of arbitrary is pulling off the top 10 highest-rated items for a user when generating item-based recommendations or the top 10 most similar users when finding user-based recommendations. That seems like kind of an arbitrary cut off. Maybe it would be better if instead of taking the top-k sources for recommendation candidates, we just used any source above some given quality threshold. For example, maybe any item a user rated higher than four stars should generate item-based recommendation candidates no matter how many or how few of them there may be. Or any user that has a cosine similarity greater than 0.95 should be used to generate candidates in the user-based recommendations. This is a pretty easy change to make, but if you're new to Python, it's a good learning experience. Go give it a try…

Contents