Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Collaborative Deep Learning for Recommender Systems Wang+, KDD’15 #221

Open
AkihikoWatanabe opened this issue Jan 11, 2018 · 2 comments
Open

Comments

@AkihikoWatanabe
Copy link
Owner

https://arxiv.org/abs/1409.2944

@AkihikoWatanabe
Copy link
Owner Author

AkihikoWatanabe commented Jan 11, 2018

Rating Matrixからuserとitemのlatent vectorを学習する際に、Stacked Denoising Auto Encoder(SDAE)によるitemのembeddingを活用する話。
Collaborative FilteringとContents-based Filteringのハイブリッド手法。
Collaborative FilteringにおいてDeepなモデルを活用する初期の研究。

通常はuser vectorとitem vectorの内積の値が対応するratingを再現できるように目的関数が設計されるが、そこにitem vectorとSDAEによるitemのEmbeddingが近くなるような項(3項目)、SDAEのエラー(4項目)を追加する。

(3項目の意義について、解説ブログより)アイテム i に関する潜在表現 vi は学習データに登場するものについては推定できるけれど,未知のものについては推定できない.そこでSDAEの中間層の結果を「推定したvi」として「真の」 vi にできる限り近づける,というのがこの項の気持ち

cite-ulikeデータによる論文推薦、Netflixデータによる映画推薦で評価した結果、ベースライン(Collective Matrix Factorization #222 , SVDFeature #223 , DeepMusic #224 , Collaborative Topic Regresison #226 )をoutperform。

image

image

image

image

image

@AkihikoWatanabe
Copy link
Owner Author

解説ブログ:http://d.hatena.ne.jp/repose/20150531/1433004688

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant