Testing exchangeability for transfer decision
Authors: 
Shuang Zhou, Evgueni Smirnov, Gijs Schoenmakers, Kurt Driessens, Ralf PeetersAuthors Info & Claims
Pages 64 - 71
Abstract
A novel solution to the problem whether to transfer based on exchangeablity test.Statistically testing if the source data is generated from the target distribution.The test is non-parametric and distribution free.Empirically justified the proposed test is effective for predicting transfer result. This paper introduces a non-parametric test to decide whether to transfer data from a source domain to a target domain to improve the generalization performance of predictive models on the target domain. The test is based on the conformal prediction framework: it statistically tests whether the target and source data are generated from the same distribution under the exchangeability assumption. The experiments show that the test is capable of outperforming existing methods when it decides on instance transfer.
References
[1]
S. Al-Stouhi, C.K. Reddy, Adaptive boosting for transfer learning using dynamic updates, Springer, 2011.
[2]
D.J. Aldous, Exchangeability and related topics, Springer, 1985.
[3]
V. Balasubramanian, S.-S. Ho, V. Vovk, Conformal prediction for reliable machine learning: Theory, adaptations and applications, Elsevier, 2014.
[4]
L. Breiman, Random forests, Mach Learn, 45 (2001) 5-32.
[5]
A. Church, On the concept of a random sequence, Bull Am Math Soc, 46 (1940) 130-135.
[6]
W. Dai, G.-R. Xue, Q. Yang, Y. Yu, Transferring naive bayes classifiers for text classification, 2007.
[7]
W. Dai, Q. Yang, G.-R. xue, Y. Yu, Boosting for transfer learning, ACM, 2007.
[8]
R. De Maesschalck, D. Jouan-Rimbaud, D.L. Massart, The mahalanobis distance, Chemometr Intell Lab Syst, 50 (2000) 1-18.
[9]
T. Fawcett, An introduction to ROC analysis, Pattern Recognit Lett, 27 (2006) 861-874.
[10]
V. Fedorova, A.J. Gammerman, I. Nouretdinov, V. Vovk, Plug-in martingales for testing exchangeability on-line, 2012.
[11]
T. Kamishima, M. Hamasaki, S. Akaho, Trbagg: a simple transfer learning method and its application to personalization in collaborative tagging, IEEE, 2009.
[12]
D. Kifer, S. Ben-David, J. Gehrke, Detecting change in data streams, VLDB Endowment, 2004.
[13]
R. Kohavi, G.H. John, Wrappers for feature subset selection, Artif Intell, 97 (1997) 273-324.
[14]
S. Kullback, R.A. Leible, On information and sufficiency, Ann Math Stat, 22 (1951) 79-86.
[15]
E.L. Lehmann, H.J. DAbrera, Nonparametrics: statistical methods based on ranks, Springer New York, 2006.
[16]
D. Lin, X. An, J. Zhang, Double-bootstrapping source data selection for instance-based transfer learning, Pattern Recognit Lett, 34 (2013) 1279-1285.
[17]
Y. Mansour, M. Mohri, A. Rostamizadeh, Domain adaptation: learning bounds and algorithms, 2009.
[18]
P. Martin-Lf, The definition of random sequences, Inf Control, 9 (1966) 602-619.
[19]
S.J. Pan, Q. Yang, A survey on transfer learning, IEEE Trans Knowl Data Eng, 22 (2010) 1345-1359.
[20]
G. Shafer, V. Vovk, A tutorial on conformal prediction, J Mach Learn Res, 9 (2008) 371-421.
[21]
B. Tan, Y. Song, E. Zhong, Q. Yang, Transitive transfer learning, ACM, 2015.
[22]
L. Torrey, J. Shavlik, Transfer learning, 2009.
[23]
Vladimir Vovk, Transductive conformal prediction, Int J Artif Intell Tools, 24 (2015) 1560001.
[24]
V. Vovk, A. Gammerman, C. Saunders, Machine-learning applications of algorithmic randomness, 1999.
[25]
V. Vovk, A. Gammerman, G. Shafer, Algorithmic learning in a random world, Springer, 2005.
[26]
E. Zhong, W. Fan, Q. Yang, O. Verscheure, J. Ren, Cross validation framework to choose amongst models and datasets for transfer learning, Springer, 2010.
[27]
S. Zhou, E. Smirnov, R. Peeters, Conformal region classification with instance-transfer boosting, Int J Artif Intell Tools, 24 (2015) 1560002.
- Testing exchangeability for transfer decision
Recommendations
Partially collapsed parallel Gibbs sampler for Dirichlet process mixture models
Proposed representation for DP is ideally suited for distributed computing.Proposed sampler offers advantages in terms of both scalability and run-time.Proposed sampler outperforms existing parallel samplers for Dirichlet process mixtures in terms of ...
Structural network inference from time-series data using a generative model and transfer entropy
Highlights- We concentrate on the problem of describing the directed flow of information between nodes based on transfer entropy.
AbstractIn this paper we develop a novel framework for inferring a generative model of network structure representing the causal relations between data for a set of objects characterized in terms of time series. To do this we make use of ...
Improved domain adaptive object detector via adversarial feature learning
AbstractUnsupervised domain adaptation is a promising solution to adapt models to new domains or environments at no additional annotation cost. Such adaptation strategy has been investigated for detection to cope with the domain shift along ...
Graphical abstractDisplay Omitted
Highlights- The proposed architecture has the advantage of performing detection in the target domain at no additional annotation.
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
Copyright © Elsevier B.V.
Publisher
Elsevier Science Inc.
United States
Publication History
Published: 01 March 2017
Author Tags
Qualifiers
- Research-article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 11 Dec 2024
Other Metrics
Citations
View Options
View options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in