Collaborative Topic Modeling for Recommending Scientific Articles - 편집 역사

Admin: /* references */

2017-07-05T16:22:24Z

references

Admin: /* references */

2017-07-05T16:21:50Z

references

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T16:18:38Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T16:18:27Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T09:46:24Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* Recommendation by Matrix Factorization */

2017-07-05T07:05:24Z

Recommendation by Matrix Factorization

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T06:11:49Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T05:26:19Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T05:25:18Z

COLLABORATIVE TOPIC REGRESSION (CTR)

Admin: /* COLLABORATIVE TOPIC REGRESSION (CTR) */

2017-07-05T05:25:01Z

COLLABORATIVE TOPIC REGRESSION (CTR)

@@ 66번째 줄: / 66번째 줄: @@
 [2] D. Agarwal and B.-C. Chen. flda: matrix factorization through latent Dirichlet allocation. In Proceedings of the third ACM international conference on Web search and data mining, WSDM ’10, pages 91–100, New York, NY, USA, 2010. ACM.
 <del>
 [3] D. Bertsekas. Nonlinear Programming. Athena Scientific, 1999.
@@ 75번째 줄: / 76번째 줄: @@
 [6] D. Blei and J. McAuliffe. Supervised topic models. In Neural Information Processing Systems, 2007.
 </del>
 [7] D. Blei, A. Ng, and M. Jordan. Latent Dirichlet allocation. Journal of Machine Learning Research, 3:993–1022, January 2003.
 [8] J. Chang, J. Boyd-Graber, S. Gerrish, C. Wang, and D. Blei. Reading tea leaves: How humans interpret topic models. In Y. Bengio, D. Schuurmans, J. Lafferty, C. K. I. Williams, and A. Culotta, editors, Advances in Neural Information Processing Systems 22, pages 288–296, 2009.
 <del>
 [9] A. Dempster, N. Laird, and D. Rubin. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B, 39:1–38, 1977.
@@ 83번째 줄: / 86번째 줄: @@
 [10] S. M. Gerrish and D. M. Blei. Predicting legislative roll calls from text. In Proceedings of the 28th Annual International Conference on Machine Learning, ICML ’11, 2011.
 </del>
 [11] J. L. Herlocker, J. A. Konstan, A. Borchers, and J. Riedl. An algorithmic framework for performing collaborative filtering. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information
 retrieval, SIGIR ’99, pages 230–237, New York, NY, USA,
@@ 91번째 줄: / 95번째 줄: @@
 [13] Y. Koren, R. Bell, and C. Volinsky. Matrix factorization techniques for recommender systems. IEEE Computer, 42(8):30–37, 2009.
 <del>
 [14] P. Melville, M. R., and R. Nagaraja. Content-boosted collaborative filtering for improved recommendations. In American Association for Artificial Intelligence, pages 187–192, 2002.
@@ 98번째 줄: / 103번째 줄: @@
 [16] R. Pan, Y. Zhou, B. Cao, N. N. Liu, R. Lukose, M. Scholz, and Q. Yang. One-class collaborative filtering. In Proceedings of the 2008 Eighth IEEE International Conference on Data Mining, pages 502–511, Washington, DC, USA, 2008. IEEE Computer Society.
 </del>
 [17] R. Salakhutdinov and A. Mnih. Bayesian probabilistic matrix factorization using Markov chain Monte Carlo. In Proceedings of the 25th International Conference on Machine learning, pages 880–887. ACM, 2008.
 [18] R. Salakhutdinov and A. Mnih. Probabilistic matrix factorization. Advances in Neural Information Processing
 Systems, 20:1257–1264, 2008.
 <del>
 [19] H. Shan and A. Banerjee. Generalized probabilistic matrix
 factorizations for collaborative filtering. In Proceedings of the 2010 IEEE International Conference on Data Mining, pages 1025–1030, Washington, DC, USA, 2010. IEEE Computer Society.
 </del>
 [20] Y. Teh, M. Jordan, M. Beal, and D. Blei. Hierarchical Dirichlet processes. Journal of the American Statistical Association, 101(476):1566–1581, 2007.
 <del>
 [21] E. Wang, D. Liu, J. Silva, D. Dunson, and L. Carin. Joint analysis of time-evolving binary matrices and associated documents. In J. Lafferty, C. K. I. Williams, J. Shawe-Taylor, R. Zemel, and A. Culotta, editors, Advances in Neural Information Processing Systems 23, pages 2370–2378. 2010.
 </del>
 [22] K. Yu, J. Lafferty, S. Zhu, and Y. Gong. Large-scale collaborative prediction using a nonparametric random effects model. In Proceedings of the 26th Annual International Conference on Machine Learning, pages 1185–1192, New York, NY, USA, 2009. ACM.

@@ 66번째 줄: / 66번째 줄: @@
 [2] D. Agarwal and B.-C. Chen. flda: matrix factorization through latent Dirichlet allocation. In Proceedings of the third ACM international conference on Web search and data mining, WSDM ’10, pages 91–100, New York, NY, USA, 2010. ACM.
+<del>
 [3] D. Bertsekas. Nonlinear Programming. Athena Scientific, 1999.
@@ 74번째 줄: / 74번째 줄: @@
 [6] D. Blei and J. McAuliffe. Supervised topic models. In Neural Information Processing Systems, 2007.
+</del>
 [7] D. Blei, A. Ng, and M. Jordan. Latent Dirichlet allocation. Journal of Machine Learning Research, 3:993–1022, January 2003.
 [8] J. Chang, J. Boyd-Graber, S. Gerrish, C. Wang, and D. Blei. Reading tea leaves: How humans interpret topic models. In Y. Bengio, D. Schuurmans, J. Lafferty, C. K. I. Williams, and A. Culotta, editors, Advances in Neural Information Processing Systems 22, pages 288–296, 2009.
+<del>
 [9] A. Dempster, N. Laird, and D. Rubin. Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B, 39:1–38, 1977.
 [10] S. M. Gerrish and D. M. Blei. Predicting legislative roll calls from text. In Proceedings of the 28th Annual International Conference on Machine Learning, ICML ’11, 2011.
+</del>
 [11] J. L. Herlocker, J. A. Konstan, A. Borchers, and J. Riedl. An algorithmic framework for performing collaborative filtering. In Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information
 retrieval, SIGIR ’99, pages 230–237, New York, NY, USA,
@@ 91번째 줄: / 91번째 줄: @@
 [13] Y. Koren, R. Bell, and C. Volinsky. Matrix factorization techniques for recommender systems. IEEE Computer, 42(8):30–37, 2009.
+<del>
 [14] P. Melville, M. R., and R. Nagaraja. Content-boosted collaborative filtering for improved recommendations. In American Association for Artificial Intelligence, pages 187–192, 2002.
@@ 97번째 줄: / 97번째 줄: @@
 [16] R. Pan, Y. Zhou, B. Cao, N. N. Liu, R. Lukose, M. Scholz, and Q. Yang. One-class collaborative filtering. In Proceedings of the 2008 Eighth IEEE International Conference on Data Mining, pages 502–511, Washington, DC, USA, 2008. IEEE Computer Society.
+</del>
 [17] R. Salakhutdinov and A. Mnih. Bayesian probabilistic matrix factorization using Markov chain Monte Carlo. In Proceedings of the 25th International Conference on Machine learning, pages 880–887. ACM, 2008.
 [18] R. Salakhutdinov and A. Mnih. Probabilistic matrix factorization. Advances in Neural Information Processing
 Systems, 20:1257–1264, 2008.
+<del>
 [19] H. Shan and A. Banerjee. Generalized probabilistic matrix
 factorizations for collaborative filtering. In Proceedings of the 2010 IEEE International Conference on Data Mining, pages 1025–1030, Washington, DC, USA, 2010. IEEE Computer Society.
+</del>
 [20] Y. Teh, M. Jordan, M. Beal, and D. Blei. Hierarchical Dirichlet processes. Journal of the American Statistical Association, 101(476):1566–1581, 2007.
+<del>
 [21] E. Wang, D. Liu, J. Silva, D. Dunson, and L. Carin. Joint analysis of time-evolving binary matrices and associated documents. In J. Lafferty, C. K. I. Williams, J. Shawe-Taylor, R. Zemel, and A. Culotta, editors, Advances in Neural Information Processing Systems 23, pages 2370–2378. 2010.
+</del>
 [22] K. Yu, J. Lafferty, S. Zhu, and Y. Gong. Large-scale collaborative prediction using a nonparametric random effects model. In Proceedings of the 26th Annual International Conference on Machine Learning, pages 1185–1192, New York, NY, USA, 2009. ACM.

@@ 53번째 줄: / 53번째 줄: @@
 \mathbf{E}[r_{ij}|u_i,θ_j,ε_j] = u_i^T (θ_j + ε_j)
 $$
-(따라서 \(u_i\)를 손보는 것은 아니다.
+(따라서 \(u_i\)를 손보는 것은 아니다. <br>
 item's latent vector \(v_j\)에 user의 선택뿐 아니라 document의 주제까지 반영하겠다는 아이디어.)

@@ 53번째 줄: / 53번째 줄: @@
 \mathbf{E}[r_{ij}|u_i,θ_j,ε_j] = u_i^T (θ_j + ε_j)
 $$
-(따라서 \(u_i\)를 손보는 것은 아니다.)
+(따라서 \(u_i\)를 손보는 것은 아니다.
 학습에 projection gradient[3]쓴다고 함.

@@ 53번째 줄: / 53번째 줄: @@
 \mathbf{E}[r_{ij}|u_i,θ_j,ε_j] = u_i^T (θ_j + ε_j)
 $$
-This is why we call the model collaborative topic regression.
+(따라서 \(u_i\)를 손보는 것은 아니다.)
 학습에 projection gradient[3]쓴다고 함.

@@ 31번째 줄: / 31번째 줄: @@
 Most successful recommendation methods are latent factor models [17, 18, 13, 1, 22], which provide better recommendation results than the neighborhood methods [11, 13]
-기본 아이디어는 user나 item이 latent factor들로 이루어진 latent vector로 표현될 수 있다고 보는 것이다. $$\hat{r}_{ij} = u^T_i v_j$$가 된다. 이를 아래와 같이 확률모델로 가정해볼 수도 있다.
+기본 아이디어는 user나 item이 [[Latent factor models for CF| latent factor]]들로 이루어진 latent vector로 표현될 수 있다고 보는 것이다. $$\hat{r}_{ij} = u^T_i v_j$$가 된다. 이를 아래와 같이 확률모델로 가정해볼 수도 있다.
 $$u_i \sim \mathcal{N} (0, \lambda_u^{-1} I_K) \\
 v_j \sim \mathcal{N} (0, \lambda_v^{-1} I_K) \\

← 이전 판		2017년 7월 5일 (수) 06:11 판
56번째 줄:		56번째 줄:

	학습에 projection gradient[3]쓴다고 함.		학습에 projection gradient[3]쓴다고 함.
−
−	coordinate ascent라는 말도 자주 나옴. 예를들어 평면상 정의되는 2-dim function이 있으면 x축에 대해 조금 진행하고, 다음에 y축에 대해 진행하고, 다시 x축에 대해 진행하고… 이런 식으로 최적점을 찾아 나가는 방법. coordinate ascent를 검색하면 coordinate descent가 나와서 둘이 똑같은건가 보다 하고 있음. [https://en.wikipedia.org/wiki/Coordinate_descent 위키]에 보면 limitations에 [https://en.wikipedia.org/wiki/Coordinate_descent#/media/File:Nonsmooth_coordinate_descent.svg 이 방법으로는 최적점으로 진행할 수 없는 케이스]가 나오는데 기가 막히다. ㅎㅎㅎ (gradient descent는 말 그대로 \(\nabla\)를 구하는거라 coordinate방향으로만 진행하지 않는다.)

	===Learning the parameters===		===Learning the parameters===

← 이전 판		2017년 7월 5일 (수) 05:26 판
58번째 줄:		58번째 줄:

	coordinate ascent라는 말도 자주 나옴. 예를들어 평면상 정의되는 2-dim function이 있으면 x축에 대해 조금 진행하고, 다음에 y축에 대해 진행하고, 다시 x축에 대해 진행하고… 이런 식으로 최적점을 찾아 나가는 방법. coordinate ascent를 검색하면 coordinate descent가 나와서 둘이 똑같은건가 보다 하고 있음. [https://en.wikipedia.org/wiki/Coordinate_descent 위키]에 보면 limitations에 [https://en.wikipedia.org/wiki/Coordinate_descent#/media/File:Nonsmooth_coordinate_descent.svg 이 방법으로는 최적점으로 진행할 수 없는 케이스]가 나오는데 기가 막히다. ㅎㅎㅎ (gradient descent는 말 그대로 \(\nabla\)를 구하는거라 coordinate방향으로만 진행하지 않는다.)		coordinate ascent라는 말도 자주 나옴. 예를들어 평면상 정의되는 2-dim function이 있으면 x축에 대해 조금 진행하고, 다음에 y축에 대해 진행하고, 다시 x축에 대해 진행하고… 이런 식으로 최적점을 찾아 나가는 방법. coordinate ascent를 검색하면 coordinate descent가 나와서 둘이 똑같은건가 보다 하고 있음. [https://en.wikipedia.org/wiki/Coordinate_descent 위키]에 보면 limitations에 [https://en.wikipedia.org/wiki/Coordinate_descent#/media/File:Nonsmooth_coordinate_descent.svg 이 방법으로는 최적점으로 진행할 수 없는 케이스]가 나오는데 기가 막히다. ㅎㅎㅎ (gradient descent는 말 그대로 \(\nabla\)를 구하는거라 coordinate방향으로만 진행하지 않는다.)
		+
		+	===Learning the parameters===
		+
		+	===Prediction===

	==references==		==references==

@@ 46번째 줄: / 46번째 줄: @@
 [12]를 읽어야 이해할 수 있는 부분이 많은듯.(특히 parameter 학습과정)
-위에서 구한 \(\theta\)를 아래에 넣음.
+위(Probabilistic Topic Models)에서 구한 \(\theta\)를 아래에 넣음.
 $$v_j = ε_j + θ_j$$