1. 1 Transfer Learning for Link Prediction Qiang Yang, 楊強，香港科大 Hong Kong University of Science and Technology Hong Kong, China http://www.cse.ust.hk/~qyang

2. We often find it easier … 2

3. 3

4. 4

5. 5

6. Transfer Learning? 迁移学习 … People often transfer knowledge to novel situations Chess  Checkers C++  Java Physics  Computer Science 6 Transfer Learning: The ability of a system to recognize and apply knowledge and skills learned in previous tasks to novel tasks (or new domains)

7. 7 But, direct application will not work Machine Learning: Training and future (test) data follow the same distribution, and are in same feature space Machine Learning: Training and future (test) data follow the same distribution, and are in same feature space

8. When distributions are different Classification Accuracy (+, -) Training: 90% Testing: 60% 8

9. When features and distributions are different Classification Accuracy (+, -) Training: 90% Testing: 10% 9

10. Training Cross-Domain Sentiment Classification Classifier Test DVD Classifier Test Electronics 82.55% 71.85% DVD

11. 11 Wireless sensor networks Different time periods, devices or space Device, Space, or Time Night time Day time Device 1 Device 2 When distributions are different

12. When Features are different Heterogeneous: different feature spaces 12 The apple is the pomaceous fruit of the apple tree, species Malus domestica in the rose family Rosaceae... Banana is the common name for a type of fruit and also the herbaceous plants of the genus Musa which produce this commonly eaten fruit... Training: TextFuture: Images Apples Bananas

13. Transfer Learning: Source Domains Learning InputOutput Source Domains 13 Source DomainTarget Domain Training DataLabeled/Unlabeled Test DataUnlabeled

14. Transfer Learning Multi-task Learning Transductive Transfer Learning Unsupervised Transfer Learning Inductive Transfer Learning Domain Adaptation Sample Selection Bias /Covariance Shift Self-taught Learning Labeled data are available in a target domain Labeled data are available only in a source domain No labeled data in both source and target domain No labeled data in a source domain Labeled data are available in a source domain Case 1 Case 2 Source and target tasks are learnt simultaneously Assumption: different domains but single task Assumption: single domain and single task An overview of various settings of transfer learning 目標數據 源數據

15. Talk Outline Transfer Learning: A quick introduction Link prediction and collaborative filtering problems Transfer Learning for Sparsity Problem Codebook Method CST Method Collective Link Prediction Method Conclusions and Future Works 15

16. Network Data Social Networks Facebook, Twitter, Live messenger, etc. Information Networks The Web (1.0 with hyperlinks and 2.0 with tags) Citation network, Wikipedia, etc. Biological Networks Gene network, etc. Other Networks 16

17. A Real World Study [Leskovec-Horvitz WWW ‘08] Who talks to whom on MSN messenger Network: 240M nodes, 1.3 billion edges Conclusions: Average path length is 6.6 90% of nodes is reachable <8 steps 17

18. Network Structures: Links Global network structures Small world Long tail distributions Local network structures Relational learning (relation==link) Link prediction (e.g. recommendation) Group (Cluster) structures 18

19. Global Network Structures Small world Six degrees of separation [Milgram 60s] Clustering and communities Evolutionary patterns Long tail distributions Personalized recommendation: Amazon Also brings a challenge in recommendation and relational learning: sparseness 19

20. Local Network Structures Link Prediction A form of Statistical Relational Learning Object classification: predict category of an object based on its attributes and links Is this person a student? Link classification: predict type of a link Are they co-authors? Link existence: predict whether a link exists or not Does he like this book? Link cardinality estimation: predict the number of links of a node An Introduction to statistical relational learning by Taskar and Getoor 20 ? ? ? ? ? ?

21. Link Prediction Task: predict missing links in a network Main challenge: Network Data Sparsity 21

22. Product Recommendation (Amazon.com) 22

24. Examples: Collaborative Filtering Training Data: Dense Rating Matrix Density >=2.0%, CF model Test Data: Rating Matrix RMSE:0.8721 Ideal Setting 131 53 312 234 44 ? ? ? ? ??

25. Examples: Collaborative Filtering Training Data: Sparse Rating Matrix Density <=0.6%, CF model Test Data: Rating Matrix RMSE:0.9748 Realistic Setting 1 5 3 4 4 ? ? ? ? ?? 10%

26. Transfer Learning for Collaborative Filtering? 26 IMDB Database Amazon.com 26

27. Codebook Transfer Bin Li, Qiang Yang, Xiangyang Xue. Can Movies and Books Collaborate? Cross-Domain Collaborative Filtering for Sparsity Reduction. In Proceedings of the Twenty-First International Joint Conference on Artificial Intelligence (IJCAI '09), Pasadena, CA, USA, July 11-17, 2009.

37. Limitations of Codebook Transfer Same rating range Source and target data must have the same range of ratings [1, 5] Homogenous dimensions User and item dimensions must be similar In reality Range of ratings can be 0/1 or [1,5] User and item dimensions may be very different 37

38. Coordinate System Transfer Weike Pan, Evan Xiang, Nathan Liu and Qiang Yang. Transfer Learning in Collaborative Filtering for Sparsity Reduction. In Proceedings of the 24th AAAI Conference on Artificial Intelligence (AAAI-10). Atlanta, Georgia, USA. July 11-15, 2010.

39. Types of User Feedback Explicit feedback: express preferences explicitly 1-5: Rating ?: Missing Implicit feedback: express preferences implicitly 1: Observed browsing/purchasing 0: Unobserved browsing/purchasing

40. Target domain R: explicit ratings One auxiliary domain (two or more data sources) : user u has implicitly browsed/purchased item j

41. Problem Formulation One target domain, n users and m items. One auxiliary domain (two data sources) user side:, shares n common users with R. item side:, shares m common items with R. Goal: predict the missing values in R.

42. Our Solution: Coordinate System Transfer Step 1: Coordinate System Construction ( ) Step 2: Coordinate System Adaptation

43. Step 1: Coordinate System Construction Sparse SVD on auxiliary data where the matrices give the top d principle coordinates. Definition (Coordinate System): A coordinate system is a matrix with columns of orthonormal bases (principle coordinates), where the columns are located in descending order according to their corresponding eigenvalues. Coordinate System Transfer

44. Step 1: Coordinate System Adaptation Adapt the discovered coordinate systems from the auxiliary domain to the target domain, The effect from the auxiliary domain Initialization: take as seed model in target domain, Regularization: Coordinate System Transfer

45. Algorithm Coordinate System Transfer

46. Data Sets and Evaluation Metrics Data sets (extracted from MovieLens and Netflix) Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), Where and are the true and predicted ratings, respectively, and is the number of test ratings. Experimental Results

47. Baselines and Parameter Settings Baselines Average Filling Latent Matrix Factorization (Bell and Koren, ICDM07) Collective Matrix Factorization (Singh and Gordon, KDD08) OptSpace (Keshavan, Montanari and Oh, NIPS10) Parameter settings Tradeoff parameters LFM: { 0.001, 0.01, 0.1, 1, 10, 100 } CMF: { 0.1, 0.5, 1, 5, 10 } CST: Latent dimensions: { 5, 10, 15 } Average results over 10 random trials are reported Experimental Results

48. Results(1/2) Observations: CST performs 99% significantly better (t-test) than all baselines in all sparsity levels, Transfer learning methods (CST, CMF) beats two non-transfer learning methods (AF, LFM). Experimental Results

49. Results(2/2) Observations: CST consistently improves as d increases, demonstrating the usefulness (avoid overfitting) using the auxiliary knowledge, The improvement (CST vs. OptSpace) is more significant with fewer observed ratings, demonstrating the effectiveness of CST in alleviating data sparsity. Experimental Results

50. Related Work (Transfer Learning) Cross-domain collaborative filtering Domain adaptation methods One-sided: Two-sided: Related Work

51. Limitation of CST and CBT Users behave differently in different domains Rating bias Users tend to rate items that they like

52. Our Solution: Collective Link Prediction (CLP) Jointly learn multiple domains together Learning the similarity of different domains consistency between domains indicates similarity.

53. Bin Cao, Nathan Liu and Qiang Yang. Transfer Learning for Collective Link Prediction in Multiple Heterogeneous Domains. In Proceedings of 27th International Conference on Machine Learning (ICML 2010), Haifa, Israel. June 2010.

54. Recommendation with Heterogeneous Domains Items span multiple, heterogeneous domains for many recommendation systems. ebay, GoogleReader, douban, etc.

55. Example: Collective Link Prediction Predicting Test Data: Rating Matrix RMSE:0.8855 1 5 3 4 4 ? ? ? ? ?? 2315 24 13 125 342 10010 01001 10110 11001 10100 Dense Auxiliary Book Rating Matrix Sparse Training Movie Rating Matrix Dense Auxiliary Movie Tagging Matrix Transfer Learning Training Different Rating Domains Different User behaviors

56. A Nonparametric Bayesian Approach Based on Gaussian process models Key part is the kernel modeling user relation as well as task relation

57. Making Prediction Similar to memory-based approach Similarity between tasks Similarity between items Mean of prediction

58. Experimental Results

59. Conclusions and Future Work Transfer Learning ( 舉一反三 ) Transfer Learning has many applications Sensor Based Location Estimation Sensor Based Activity Recognition Future: When to transfer, and when not to transfer Finding source domains on World Wide Web Heterogeneous transfer learning 59