Data-driven adaptive selection of rule quality measures for improving rule induction and filtration algorithms

This paper presents a proposal of a rule induction algorithm selecting a rule quality measure adaptively. The quality measure plays the role of an optimization criterion of the generated rules. Nine quality measures applied by the algorithm are presented and discussed in the paper. It is shown experimentally that the proposed algorithm provides us with obtaining a classifier of the best quality. During experiments, three criteria of the classifier quality were considered: overall accuracy, balanced accuracy (average accuracy of decision classes), and complexity of the classifier (understood to mean the number of induced rules). The experiments were carried out on 34 data sets coming from the UCI machine learning repository. Moreover, a proposal of four-rule filtration algorithms is presented in the paper. Their task is to limit the number of rules in the classifier. In particular, filtration influence on the classifier quality is studied.

[1]  Sebastian Widz,et al.  Is It Important Which Rough-Set-Based Classifier Extraction and Voting Criteria Are Applied Together? , 2010, RSCTC.

[2]  Salvatore Greco,et al.  Mining Pareto-optimal rules with respect to support and confirmation or support and anti-support , 2007, Eng. Appl. Artif. Intell..

[3]  Peter A. Flach,et al.  Subgroup Discovery with CN2-SD , 2004, J. Mach. Learn. Res..

[4]  Gerd Stumme,et al.  Efficient Mining of Association Rules Based on Formal Concept Analysis , 2005, Formal Concept Analysis.

[5]  Sigal Sahar,et al.  Interestingness Measures - On Determining What Is Interesting , 2005, Data Mining and Knowledge Discovery Handbook.

[6]  Josef Tkadlec,et al.  Rule quality for multiple-rule classifier: Empirical expertise and theoretical methodology , 2003, Intell. Data Anal..

[7]  E. Eells,et al.  Symmetries and Asymmetries in Evidential Support , 2002 .

[8]  Johannes Fürnkranz,et al.  ROC ‘n’ Rule Learning—Towards a Better Understanding of Covering Algorithms , 2005, Machine Learning.

[9]  Andrzej Skowron,et al.  Transactions on Rough Sets III , 2005, Trans. Rough Sets.

[10]  Janusz Zalewski,et al.  Rough sets: Theoretical aspects of reasoning about data , 1996 .

[11]  Zdzisław Pawlak,et al.  Can Bayesian confirmation measures be useful for rough set decision rules? , 2004, Eng. Appl. Artif. Intell..

[12]  Howard J. Hamilton,et al.  Knowledge discovery and measures of interest , 2001 .

[13]  James M. Joyce.  The Foundations of Causal Decision Theory by James M. Joyce , 1999 .

[14]  Peter A. Flach,et al.  Rule Evaluation Measures: A Unifying View , 1999, ILP.

[15]  JOHANNES FÜRNKRANZ,et al.  Separate-and-Conquer Rule Learning , 1999, Artificial Intelligence Review.

[16]  Rakesh Agarwal,et al.  Fast Algorithms for Mining Association Rules , 1994, VLDB 1994.

[17]  Frank Höppner,et al.  Association Rules , 2005, Data Mining and Knowledge Discovery Handbook.

[18]  Johannes Fürnkranz,et al.  A review and comparison of strategies for handling missing values in separate-and-conquer rule learning , 2011, Journal of Intelligent Information Systems.

[19]  William W. Cohen.  Fast Effective Rule Induction , 1995, ICML.

[20]  James M. Joyce.  The Foundations of Causal Decision Theory , 1999 .

[21]  Dominik Slezak,et al.  Rough Sets and Functional Dependencies in Data: Foundations of Association Reducts , 2009, Trans. Comput. Sci..

[22]  Kenneth McGarry,et al.  A survey of interestingness measures for knowledge discovery , 2005, The Knowledge Engineering Review.

[23]  Wlodzislaw Duch,et al.  A new methodology of extraction, optimization and application of crisp and fuzzy logical rules , 2001, IEEE Trans. Neural Networks.

[24]  Jerzy W. Grzymala-Busse,et al.  Data mining based on rough sets , 2003 .

[25]  อนิรุธ สืบสิงห์,et al.  Data Mining Practical Machine Learning Tools and Techniques , 2014 .

[26]  Tom Fawcett.  PRIE: a system for generating rulelists to maximize ROC performance , 2008, Data Mining and Knowledge Discovery.

[27]  Jaideep Srivastava,et al.  Selecting the right interestingness measure for association patterns , 2002, KDD.

[28]  Dominik Slezak,et al.  Rough Sets and Bayes Factor , 2005, Trans. Rough Sets.

[29]  Sebastian Widz,et al.  Rough-Set-Inspired Feature Subset Selection, Classifier Construction, and Rule Aggregation , 2011, RSKT.

[30]  Ryszard S. Michalski,et al.  Discovering Classification Rules Using variable-Valued Logic System VL1 , 1973, IJCAI.

[31]  Daniel Vanderpooten,et al.  Induction of decision rules in classification and discovery-oriented perspectives , 2001, Int. J. Intell. Syst..

[32]  Gregory Piatetsky-Shapiro,et al.  Discovery, Analysis, and Presentation of Strong Rules , 1991, Knowledge Discovery in Databases.

[33]  Łukasz Wróbel,et al.  Application of Rule Induction Algorithms for Analysis of Data Collected by Seismic Hazard Monitoring Systems in Coal Mines , 2010 .

[34]  Marzena Kryszkiewicz,et al.  Representative Association Rules and Minimum Condition Maximum Consequence Association Rules , 1998, PKDD.

[35]  Yiyu Yao,et al.  Micro and macro evaluation of classification rules , 2008, 2008 7th IEEE International Conference on Cognitive Informatics.

[36]  Johannes Fürnkranz,et al.  On the quest for optimal rule learning heuristics , 2010, Machine Learning.

[37]  Howard J. Hamilton,et al.  Interestingness measures for data mining: A survey , 2006, CSUR.

[38]  Marek Sikora,et al.  Decision Rule-Based Data Models Using TRS and NetTRS - Methods and Algorithms , 2010, Trans. Rough Sets.

[39]  Bing Liu,et al.  Generating Classification Rules According to User's Existing Knowledge , 2001, SDM.

[40]  Nada Lavrac,et al.  Expert-Guided Subgroup Discovery: Methodology and Application , 2011, J. Artif. Intell. Res..

[41]  Marek Sikora,et al.  Rule Quality Measures in Creation and Reduction of Data Rule Models , 2006, RSCTC.

[42]  Rajjan Shinghal,et al.  Evaluating the Interestingness of Characteristic Rules , 1996, KDD.

[43]  Marek Sikora,et al.  Data-Driven Adaptive Selection of Rules Quality Measures for Improving the Rules Induction Algorithm , 2011, RSFDGrC.

[44]  Hui Xiong,et al.  Exploiting a support-based upper bound of Pearson's correlation coefficient for efficiently identifying strongly correlated pairs , 2004, KDD.

[45]  Yiyu Yao,et al.  An Analysis of Quantitative Measures Associated with Rules , 1999, PAKDD.

[46]  S Dzeroski,et al.  Rule induction and instance-based learning applied in medical diagnosis. , 1996, Technology and health care : official journal of the European Society for Engineering and Medicine.

[47]  Shusaku Tsumoto,et al.  Evaluation of rule interestingness measures in medical knowledge discovery in databases , 2007, Artif. Intell. Medicine.

[48]  Dominik Slezak,et al.  The Lower and the Upper Systems of Rules in Tables with Missing Values , 2010, FGIT-DTA/BSBT.

[49]  Nick Cercone,et al.  Rule Quality Measures for Rule Induction Systems: Description and Evaluation , 2001, Comput. Intell..

[50]  Kenneth A. Kaufman,et al.  LEARNING IN AN INCONSISTENT WORLD RULE SELECTION IN STAR / AQ 18 , 1999 .

[51]  Marek Sikora,et al.  Induction and selection of the most interesting Gene Ontology based multiattribute rules for descriptions of gene groups , 2011, Pattern Recognit. Lett..

[52]  Branko Kavsek,et al.  APRIORI-SD: ADAPTING ASSOCIATION RULE LEARNING TO SUBGROUP DISCOVERY , 2006, IDA.

[53]  Geoffrey I. Webb,et al.  Generality Is Predictive of Prediction Accuracy , 2006, Selected Papers from AusDM.

[54]  Izabela Szczech,et al.  Multicriteria Attractiveness Evaluation of Decision and Association Rules , 2009, Trans. Rough Sets.

[55]  Marek Sikora,et al.  Induction and pruning of classification rules for prediction of microseismic hazards in coal mines , 2011, Expert Syst. Appl..