riccardopoiani recsys-competition issues

riccardopoiani / recsys-competition

Reccomender systems project

MIT License

1 stars 1 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

IALS with side info and different alpha for urm and icm

#61 tangtang95 opened 4 years ago
0
Discretize price and asset by giving also small values to neighbour category

#60 tangtang95 closed 4 years ago
0
Future directions

#59 riccardopoiani opened 4 years ago
0
Peculiarities and behavior to be understood

#58 riccardopoiani opened 4 years ago
0
Exploration ideas

#57 riccardopoiani closed 4 years ago
0
Hybrid reranking recommender

#56 tangtang95 closed 4 years ago
0
Apply FFM on smaller subset of dataset

#55 tangtang95 closed 4 years ago
0
P3alpha with user side info

#54 tangtang95 closed 4 years ago
0
Feature weighting

#53 tangtang95 closed 4 years ago
0
Use surprise library

#52 tangtang95 closed 4 years ago
0
On the way to the first deadline

#51 riccardopoiani closed 4 years ago
1
Boost userSimilarityRecommender

#50 tangtang95 closed 4 years ago
1
Check carefully that UCM returned by data splitter is correct

#49 tangtang95 closed 4 years ago
0
Try tuning by changing parameters of okapi BM25

#48 tangtang95 closed 4 years ago
0
Try to apply TF_IDF on URM_train.T

#47 tangtang95 closed 4 years ago
2
Re-tuning algorithms that has the new parameter “interactions_feature_weighting"

#46 tangtang95 closed 4 years ago
0
Try to apply feature weighting also on URM

#45 tangtang95 closed 4 years ago
0
Feature engineering

#44 tangtang95 opened 4 years ago
1
Improve performances on high tale of user profiles / users with age = 4

#43 riccardopoiani opened 4 years ago
0
XGBoost

#42 riccardopoiani closed 4 years ago
0
On the way to the third submission

#41 riccardopoiani closed 4 years ago
0
Refactor custom recommender names

#40 riccardopoiani closed 4 years ago
0
Bagging

#39 tangtang95 closed 4 years ago
0
Fix format_URM_slice

#38 tangtang95 closed 4 years ago
0
CBF tuning with ICM_all euclidean

#37 tangtang95 closed 4 years ago
0
Try less coarse plot for the GROUP hybrid

#36 riccardopoiani closed 4 years ago
0
Check user cold performances

#35 riccardopoiani closed 4 years ago
0
FM

#34 riccardopoiani closed 4 years ago
0
SSLim tuning

#33 riccardopoiani closed 4 years ago
0
Model merge

#32 riccardopoiani closed 4 years ago
0
New lib

#31 riccardopoiani closed 4 years ago
1
Boost speed of plots

#30 tangtang95 closed 4 years ago
0
Boost hybrid tuning

#29 tangtang95 closed 4 years ago
1
New hybrid

#28 riccardopoiani opened 4 years ago
0
Fix UserItemCBF_CF_DemographicRecommender

#27 tangtang95 closed 4 years ago
0
Re-evaluate all recommenders in best models for only warm users

#26 tangtang95 closed 4 years ago
0
Try LightFM

#25 tangtang95 closed 4 years ago
1
Add a parameters file in plot graphics

#24 tangtang95 closed 4 years ago
0
Explore cold items

#23 riccardopoiani opened 4 years ago
0
Add new demographics

#22 riccardopoiani closed 4 years ago
0
Fix cross validation for Hybrid recommenders

#21 riccardopoiani closed 4 years ago
0
Improving cold start performances

#20 riccardopoiani closed 4 years ago
0
Hybrid Demographic + Any recommender

#19 tangtang95 closed 4 years ago
0
User CBF

#18 tangtang95 closed 4 years ago
0
Tuning MF

#17 tangtang95 closed 4 years ago
1
Mega Refactor

#16 tangtang95 closed 4 years ago
0
Clustering of users

#15 tangtang95 closed 4 years ago
0
Hexplot recommendations

#14 tangtang95 opened 4 years ago
0
Hybrid ICM: mix model learned with ICM categorical and ICM numerical

#13 tangtang95 closed 4 years ago
0
ICM: Feature augmentation

#12 tangtang95 closed 4 years ago
0