Recommender systems and Machine Learning

Question

According to Mitchell:

“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T , as measured by P , improves with experience E” (Mitchell, 1997) .

If we put this in terms of recommender systems:

Task: recommend products

Experience: includes the experience of observing a set of examples encoded in a ratings matrix

Performance: metric i.e. rmse between real and predicted ratings

Is my approach okay?

Any suggestions are welcomed!

Answer 1

• For experience:

observing a set of examples encoded in a rating matrix is good, before doing that perform hybrid filtering(content-based filtering + collaborative), from its results get insights and do rating matrix if needed.

• Performance:

  use rmse metric for better understanding of accuracy.