search

Chinmayrane16 / DeepRecommender

Training Deep AutoEncoders for Collaborative Filtering
MIT License
22 stars 3 forks source link

Generating Top-N Recommendations for the Recommendation System #1

Open PurushothamanSrikanth opened 3 years ago

PurushothamanSrikanth commented 3 years ago

@Chinmayrane16 Can we have Top-N recommendation as a new feature to the existing one? We can generate Top-N recommendations from the rating prediction matrix.

This is a feature request 😇

Chinmayrane16 commented 3 years ago

@PurushothamanSrikanth Yes, it is possible and fairly simple. At inference, we can get the Top-N indexes (movies) from the user's rating vector with the maximum scores.

Do you wish to proceed and make a PR? ;)

PurushothamanSrikanth commented 3 years ago

@PurushothamanSrikanth Yes, it is possible and fairly simple. At inference, we can get the Top-N indexes (movies) from the user's rating vector with the maximum scores.

Do you wish to proceed and make a PR? ;)

PurushothamanSrikanth commented 3 years ago

@PurushothamanSrikanth Yes, it is possible and fairly simple. At inference, we can get the Top-N indexes (movies) from the user's rating vector with the maximum scores. Do you wish to proceed and make a PR? ;)

  • [ ] That's great. I'm trying that, but I'm not able to do it properly.

@Chinmayrane16 Are u talking about out matrix in [Training_Deep_AE.ipynb] ?(https://github.com/Chinmayrane16/DeepRecommender/blob/master/Training_Deep_AE.ipynb)

Untitled1

Chinmayrane16 commented 3 years ago

@PurushothamanSrikanth Yeah, you could use that. Choose a particular index (user) and take its rating vector and select Top-N maximum indexes. Map those indexes back to Movie names. :)

PurushothamanSrikanth commented 3 years ago

@PurushothamanSrikanth Yeah, you could use that. Choose a particular index (user) and take its rating vector and select Top-N maximum indexes.

Map those indexes back to Movie names. :)

Will give that a try