Correcting evaluation bias of relational classifiers with network cross validation
Abstract
Recently, a number of modeling techniques have been developed for data mining and machine learning in relational and network domains where the instances are not independent and identically distributed (i.i.d.). These methods specifically exploit the statistical dependencies among instances in order to improve classification accuracy. However, there has been little focus on how these same dependencies affect our ability to draw accurate conclusions about the performance of the models. More specifically, the complex link structure and attribute dependencies in relational data violate the assumptions of many conventional statistical tests and make it difficult to use these tests to assess the models in an unbiased manner. In this work, we examine the task of within-network classification and the question of whether two algorithms will learn models that will result in significantly different levels of performance. We show that the commonly used form of evaluation (paired t-test on overlapping network samples) can result in an unacceptable level of Type I error. Furthermore, we show that Type I error increases as (1) the correlation among instances increases and (2) the size of the evaluation set increases (i.e., the proportion of labeled nodes in the network decreases). Lastly, we propose a method for networkmore »
- Authors:
-
- Purdue Univ., West Lafayette, IN (United States)
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Rutgers Univ., Piscataway, NJ (United States)
- Publication Date:
- Research Org.:
- Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1343835
- Report Number(s):
- LLNL-JRNL-455699
Journal ID: ISSN 0219-1377
- Grant/Contract Number:
- AC52-07NA27344
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Knowledge and Information Systems
- Additional Journal Information:
- Journal Volume: 30; Journal Issue: 1; Journal ID: ISSN 0219-1377
- Publisher:
- Springer
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 99 GENERAL AND MISCELLANEOUS; 97 MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; relational learning; collective classification; statistical tests; methodology
Citation Formats
Neville, Jennifer, Gallagher, Brian, Eliassi-Rad, Tina, and Wang, Tao. Correcting evaluation bias of relational classifiers with network cross validation. United States: N. p., 2011.
Web. doi:10.1007/s10115-010-0373-1.
Neville, Jennifer, Gallagher, Brian, Eliassi-Rad, Tina, & Wang, Tao. Correcting evaluation bias of relational classifiers with network cross validation. United States. https://doi.org/10.1007/s10115-010-0373-1
Neville, Jennifer, Gallagher, Brian, Eliassi-Rad, Tina, and Wang, Tao. Tue .
"Correcting evaluation bias of relational classifiers with network cross validation". United States. https://doi.org/10.1007/s10115-010-0373-1. https://www.osti.gov/servlets/purl/1343835.
@article{osti_1343835,
title = {Correcting evaluation bias of relational classifiers with network cross validation},
author = {Neville, Jennifer and Gallagher, Brian and Eliassi-Rad, Tina and Wang, Tao},
abstractNote = {Recently, a number of modeling techniques have been developed for data mining and machine learning in relational and network domains where the instances are not independent and identically distributed (i.i.d.). These methods specifically exploit the statistical dependencies among instances in order to improve classification accuracy. However, there has been little focus on how these same dependencies affect our ability to draw accurate conclusions about the performance of the models. More specifically, the complex link structure and attribute dependencies in relational data violate the assumptions of many conventional statistical tests and make it difficult to use these tests to assess the models in an unbiased manner. In this work, we examine the task of within-network classification and the question of whether two algorithms will learn models that will result in significantly different levels of performance. We show that the commonly used form of evaluation (paired t-test on overlapping network samples) can result in an unacceptable level of Type I error. Furthermore, we show that Type I error increases as (1) the correlation among instances increases and (2) the size of the evaluation set increases (i.e., the proportion of labeled nodes in the network decreases). Lastly, we propose a method for network cross-validation that combined with paired t-tests produces more acceptable levels of Type I error while still providing reasonable levels of statistical power (i.e., 1–Type II error).},
doi = {10.1007/s10115-010-0373-1},
journal = {Knowledge and Information Systems},
number = 1,
volume = 30,
place = {United States},
year = {Tue Jan 04 00:00:00 EST 2011},
month = {Tue Jan 04 00:00:00 EST 2011}
}
Web of Science
Works referenced in this record:
Enhanced hypertext categorization using hyperlinks
journal, June 1998
- Chakrabarti, Soumen; Dom, Byron; Indyk, Piotr
- ACM SIGMOD Record, Vol. 27, Issue 2
Cautious Inference in Collective Classification
report, July 2007
- McDowell, Luke K.; Gupta, Kalyan M.; Aha, David W.
- AAAI Conference on Artificial Intelligence
Learning relational probability trees
conference, January 2003
- Neville, Jennifer; Jensen, David; Friedland, Lisa
- Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '03
Leveraging Relational Autocorrelation with Latent Group Models
conference, January 2005
- Neville, J.; Jensen, D.
- Fifth IEEE International Conference on Data Mining (ICDM'05)
Distribution-based aggregation for relational learning with identifier attributes
journal, January 2006
- Perlich, Claudia; Provost, Foster
- Machine Learning, Vol. 62, Issue 1-2
An Examination of Experimental Methodology for Classifiers of Relational Data
conference, October 2007
- Gallagher, Brian; Eliassi-Rad, Tina
- 2007 Seventh IEEE International Conference on Data Mining - Workshops (ICDM Workshops), Seventh IEEE International Conference on Data Mining Workshops (ICDMW 2007)
Exploiting Cluster-Structure to Predict the Labeling of a Graph
book, January 2008
- Herbster, Mark
- Lecture Notes in Computer Science
Using ghost edges for classification in sparsely labeled networks
conference, January 2008
- Gallagher, Brian; Tong, Hanghang; Eliassi-Rad, Tina
- Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD 08
Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms
journal, October 1998
- Dietterich, Thomas G.
- Neural Computation, Vol. 10, Issue 7
Simple estimators for relational Bayesian classifiers
conference, January 2003
- Neville, J.; Jensen, D.; Gallagher, B.
- Third IEEE International Conference on Data Mining
Why collective inference improves relational classification
conference, January 2004
- Jensen, David; Neville, Jennifer; Gallagher, Brian
- Proceedings of the 2004 ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '04
Collective Classification in Network Data
journal, September 2008
- Sen, Prithviraj; Namata, Galileo; Bilgic, Mustafa
- AI Magazine, Vol. 29, Issue 3
Evaluating Statistical Tests for Within-Network Classifiers of Relational Data
conference, December 2009
- Neville, Jennifer; Gallagher, Brian; Eliassi-Rad, Tina
- 2009 Ninth IEEE International Conference on Data Mining (ICDM)
National Longitudinal Study of Adolescent to Adult Health (Add Health), 1994-2008 [Public Use] (ICPSR 21600). Version 18
dataset, October 2017
- Harris, Kathleen Mullan; Udry, J. Richard
- Data Sharing for Demographic Research , multi-year/versions
Enhanced hypertext categorization using hyperlinks
conference, January 1998
- Chakrabarti, Soumen; Dom, Byron; Indyk, Piotr
- Proceedings of the 1998 ACM SIGMOD international conference on Management of data - SIGMOD '98
Leveraging relational autocorrelation with latent group models
conference, January 2005
- Neville, Jennifer; Jensen, David
- Proceedings of the 4th international workshop on Multi-relational mining - MRDM '05
Leveraging relational autocorrelation with latent group models
text, January 2006
- Neville, Jennifer; Jensen, David
- Schloss Dagstuhl - Leibniz-Zentrum für Informatik
Learning Probabilistic Relational Models
book, January 2001
- Getoor, Lise; Friedman, Nir; Koller, Daphne
- Relational Data Mining
Enhanced hypertext categorization using hyperlinks
conference, January 1998
- Chakrabarti, Soumen; Dom, Byron; Indyk, Piotr
- Proceedings of the 1998 ACM SIGMOD international conference on Management of data - SIGMOD '98
Approximate Statistical Tests for Comparing Supervised Classification Learning Algorithms
journal, October 1998
- Dietterich, Thomas G.
- Neural Computation, Vol. 10, Issue 7
Works referencing / citing this record: