Differences

This shows you the differences between two versions of the page.

people:bo_xu:sorec [2010/12/01 15:39]
box
people:bo_xu:sorec [2011/01/06 19:51] (current)
box
Line 88: Line 88:
==== Collaborative filtering system ==== ==== Collaborative filtering system ====
 + === Heuristic-based Methods ===
-    - **Probabilistic Matrix Factorization (PMF)** (NIPS, 2008)[[http://web.mit.edu/~rsalakhu/www/papers/nips07_pmf.pdf|download(pdf)]]------ [[http://www.wikicoursenote.com/wiki/Probabilistic_Matrix_Factorization|Introduction]]+ 
 + 
 + 
 + === Model-based Methods === 
 +    - **Learning collaborative information filters** (ICML, 1998)[[http://www.stanford.edu/~koutrika/Readings/res/Default/billsus98learning.pdf | (pdf)]] 
 +        -    **Abstract:** Model collaborative filtering task as the classification or regression problem in machine learning and Apply SVD to reduce the dimensionality. (**convert the training data,the sparse matrix of user ratings to Boolean feature vetors, resulting in a matrix filled with zeros and ones)**. 
 +        - **Explore :** 
 +    - **Empirical analysis of predictive algorithms for collaborative filtering** (UAI, 1998) [[ftp://ftp.research.microsoft.com/pub../TR/TR-98-12.pdf|pdf]] 
 +        -    **Abstract:** Cluster Models and Bayes Network model is embedded. Train the parameter by training data and predict the unknown ratings. <code> Naive Bayes: p(C = c, v_1, v_2, ... v_n)= Pr(C = c)\prod_{i=1} ^n (v_i|C=c) </code> 
 +        - **Explore :**  
 +    - **Using Probabilistic Relational Models for Collaborative Filtering** (WebKDD 1999)[[http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.40.4507&rep=rep1&type=pdf|(pdf)]] 
 +        -    **Abstract:** Apply probabilistic relational models (PRM),which is similar to [[http://en.wikipedia.org/wiki/Entity-relationship_model| Entity relational model]], to the system. 
 +        - **Explore :**  
 +    - **Eigentaste: A constant Time Collaborative Filtering Algorithm** (Information Retrival J, 2001)[[http://ml.stat.purdue.edu/vdb/joke/Papers/eigentaste.pdf|(pdf)]] 
 +        -    **Abstract:** Select gauge set (all valid users rated all items in the gauge set) and apply PCA for dimensionality reduction. Then cluster users. Classify new users to the corresponding cluster and recommend items.  
 +        - **Explore :** 1. Cluster method not fit to multi-model. 2. gauge set may hard to select 3. when gauge set is small, the recommend is not accurate. 
 +    - **Collaborative Filtering via Gaussian Probabilistic Latent Semantic Analysis** (SIGIR, 2003)[[http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.7.2476&rep=rep1&type=pdf|(pdf)]] 
 +        -    **Abstract:** Model collaborative filtering task as the classification or regression problem in machine learning and Apply SVD to reduce the dimensionality. 
 +        - **Explore :**  
 +    - **Modeling User Rating Profiles for Collaborative Filtering** (NIPS, 2003)[[http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.68.7750&rep=rep1&type=pdf|(pdf)]] 
 +        -    **Abstract:** Model collaborative filtering task as the classification or regression problem in machine learning and Apply SVD to reduce the dimensionality. 
 +        - **Explore :**  
 +    - **A Maximum Entropy Approach to Collaborative Filtring in Dynamic, Sparse, High-Dimensional Domains** (NIPS, 2002)[[http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.68.9008&rep=rep1&type=pdf|(pdf)]] 
 +        -    **Abstract:** Model collaborative filtering task as the classification or regression problem in machine learning and Apply SVD to reduce the dimensionality. 
 +        - **Explore :**  
 +    - **Clustring Methods for Collaborative Filtring**(Technical Report, 1998)[[http://reference.kfupm.edu.sa/content/c/l/clustering_methods_for_collaborative_fil_120667.pdf|(pdf)]] 
 +        -    **Abstract:** Model collaborative filtering task as the classification or regression problem in machine learning and Apply SVD to reduce the dimensionality. 
 +        - **Explore :**  
 +    - **Probabilistic Matrix Factorization (PMF)** (NIPS, 2008)[[http://web.mit.edu/~rsalakhu/www/papers/nips07_pmf.pdf|(pdf)]]------ [[http://www.wikicoursenote.com/wiki/Probabilistic_Matrix_Factorization|Introduction]]
       - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.        - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.
       - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.        - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.
-    - **Bayesian probabilistic matrix factorization using Markov chain Monte Carlo** (ICML, 2008)[[http://www.cs.utoronto.ca/~amnih/papers/bpmf.pdf|download(pdf)]]+    - **Bayesian probabilistic matrix factorization using Markov chain Monte Carlo** (ICML, 2008)[[http://www.cs.utoronto.ca/~amnih/papers/bpmf.pdf|(pdf)]]
       - **Abstract:** Present a full Bayesian treatment of the PMF model in which model capacity is controlled automatically by integrating over all model parameters and hyperparameters.        - **Abstract:** Present a full Bayesian treatment of the PMF model in which model capacity is controlled automatically by integrating over all model parameters and hyperparameters.
       - **Superiority:** higher prediction accuracy thant PMF models.          - **Superiority:** higher prediction accuracy thant PMF models.  
-     - **Sorec: social recommendation using probabilistic matrix factorization** (CIKM, 2008)[[http://www.cse.cuhk.edu.hk/~lyu/paper_pdf/Social%20Recommendation-ma.pdf|download(pdf)]]+     - **Sorec: social recommendation using probabilistic matrix factorization** (CIKM, 2008)[[http://www.cse.cuhk.edu.hk/~lyu/paper_pdf/Social%20Recommendation-ma.pdf|(pdf)]]
       - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.        - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.
       - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.        - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.
-     - **Learning to Recommend with Social Trust Ensemble** (SIGIR, 2009)[[http://www.cse.cuhk.edu.hk/~king/PUB/SIGIR2009-p203.pdf|download(pdf)]]+     - **Learning to Recommend with Social Trust Ensemble** (SIGIR, 2009)[[http://www.cse.cuhk.edu.hk/~king/PUB/SIGIR2009-p203.pdf|(pdf)]]
       - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.        - **Abstract:** PMF apply a probabilistic approach using Gaussian assumptions on the knonw data and the factor matrics to factor the matrix and pridicting the missing values.Experimental resuts show that PMF perform quite well.
       - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.        - **Superiority:** Scales linearly, performs well on the large, spase and imbalanced dataset.
Line 119: Line 148:
===== works ===== ===== works =====
 +
 +list of some papers:
 +
 +1. Relational learning via collective matrix Factorization : Ajit P.Singh
 +2. Locality Preserving Nonnegative matrix factorization dengcai
 +3. relation regularized matrix factorization, wu0jun Li
 +4. Modeling user rating Profiles for collaborative filtering
 +5. Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions
 +6. application of dimensionality reduction in recommender system-a case study
 +7. collaborative ifltering via guassian probabilistic latent semantic analysis
 +8 item based collaborative filtering recommendation algorithms
 +9. maximum likelihood estimation of intrinsic dimension
 +10 Optimization algorithms in machine learning --- stephen wright
 +11. global analytic solution for variational bayesian matrix factorization
 +12. variational bayesian approach to movie rating prediction
 +13. implicit regularization in variational bayesian matrix factorization
 +14. sparse inverse covariance estimation with the graphical lasso
 +15. matrix factorization techniques for recommender systemns
 +16probabilistic sparse matrix factorization
 +17. learning with local and global consistency
 +18
 
people/bo_xu/sorec.1291189141.txt.gz · Last modified: 2010/12/01 15:39 by box     Back to top