Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
xiaosen
b28f5500ad
|
1 year ago | |
|---|---|---|
| .idea | 1 year ago | |
| data | 1 year ago | |
| models | 1 year ago | |
| src | 1 year ago | |
| .gitignore | 1 year ago | |
| LICENSE | 1 year ago | |
| README.md | 1 year ago | |
README.md
Neural Collaborative Filtering
Pythorch Version of Neural Collaborative Filtering at WWW'17
Keypoints
- The problem that the thesis intends to solve is to recommend the item to the user based on implicit feedback.
- Applying deep learning to user-item interaction in matrix factorization
- Using a network structure that takes advantage of both dot-product (GMF) and MLP
- Use binary cross-entropy rather than MSE as loss function
- Use point-wise loss + negative sampling
Results
Best epoch 005: HR = 0.677, NDCG = 0.399
Requirements
python==3.7.7
pandas==1.0.3
numpy==1.18.1
torch==1.4.0