I implemented KNN with learning to find the best K, using cosine similarity and similarity weighting for prediction. Also the prediction is clamped to the range [1,5] since we know that ratings are in that range.

Run this program on a dataset Arrow_right

Existing runs on cberner-KNNImproved-r 1-10 of 10   Action_refresh_blue
ID Program Dataset Tuned hyper. User Updated << Status Total time Memory Error
Run #40179 cberner-KNNImproved-r testlg no soryn1 2y253d ago failed 6s 418M
Run #30316 cberner-KNNImproved-r CollaborativeFiltering-ZL no shakey 4y139d ago failed 7m5s 1523M
Run #17606 cberner-KNNImproved-r WordSegmentationData no internal 5y290d ago done 0.340
Run #17607 cberner-KNNImproved-r CollaborativeFilteringData no internal 5y290d ago done 1.71
Run #2068 cberner-KNNImproved-r cs281amovielens no internal 7y195d ago done 1s 415M 0.761
Run #1336 cberner-KNNImproved-r movielens1m no jake 7y195d ago done 1h29m 833M 0.905
Run #1312 cberner-KNNImproved-r eachmovie-1to5 no lester 7y195d ago failed 24m27s 2010M
Run #1231 cberner-KNNImproved-r eachmovie no lester 7y195d ago done 1h28m 1764M 1.20
Run #875 cberner-KNNImproved-r movielens100k no cberner 7y195d ago done 5m26s 143M 0.942
Run #874 cberner-KNNImproved-r collaborativefiltering-sample no cberner 7y195d ago done 1s 428M 0.901

Program contents Arrow_right


Must be logged in to post comments.