潜在トピックモデルに ... - latent-dynamics.net

1 2 3 D. Blei, A. Ng, M. Jordan, Latent Dirichlet Allocation, JMLR2003 4 D. Blei, J. Lafferty, Dynamic Topic Models, ICML20065 [L. Cao and L. Fei-Fei. Spatially coherent latent topic model for concurrent object segmentation and classification . ICCV2007]6 7bag-of-word W W 1w2w3w4w5w6w7w8w w8 1()( )NnnPPw== w w ()Pw9 1()() (|)NnznPPzPw z== w (|)Pw z ()Pzzw 10 1()(|)( |)NnznPPz Pw z == w (|)Pw z(|)Pz zw 11 PLSA LDA PLSA =PLSI : Probabilistic Latent Semantic analysis LDA: Latent Dirichlet Allocation PLSA+ 1() ()(|)( |)NnznPPPzPwzd == w1()(|)( |)NnznPPz Pw z == w 12 D.

24 Topic model • Probabilistic latent semantic analysis • パラメータ数を潜在トピックの導入により 削減 ユーザ毎の興味 トピック毎の流行

Loading..

Tags:

  Analysis, Talent, Semantics, Latent semantic analysis

Information

Related search queries