The impact of measurement scale and correlation structure on classification performance of inductive learning and statistical methods

Abstract This is a comparative study of inductive learning and statistical methods using the simulation approach to provide a generalizable results. The purpose of this study is to investigate the impact of measurement scale of explanatory variables on the relative performance of the statistical method (probit) and the inductive learning method (ID3) and to examine the impact of correlation structure on the classification behavior of the probit method and the ID3 method. The simulation results show that the relative classification accuracy of ID3 to probit increases as the proportion of binary variables increases in the classification model, and that the relative accuracy of ID3 to probit is higher when the covariance matrices are unequal among populations than when the covariance matrices are equal among populations. The empirical tests on ID3 reveal that the classification accuracy of ID3 is lower when the covariance matrices are unequal among populations than when the covariance matrices are equal among populations and that the classification accuracy of ID3 decreases as the correlations among explanatory variables increases.

[1]  Bay Arinze,et al.  An experimental investigation of the predictive accuracy of induction and regression , 1994 .

[2]  T. Amemiya.  QUALITATIVE RESPONSE MODELS: A SURVEY , 1981 .

[3]  V. E. Kane,et al.  Assessment of fisher and logistic linear and quadratic discrimination models , 1983 .

[4]  Ryszard S. Michalski,et al.  INDUCE.3: A Program for Learning Structural Descriptions from Examples , 1983 .

[5]  D. Moore.  Evaluation of Five Discrimination Procedures for Binary Variables , 1973 .

[6]  Bijayananda Naik,et al.  Using Rule Induction for Expert System Development: The Case of Asset Writedowns , 1994 .

[7]  Ingoo Han,et al.  An empirical investigation of some data effects on the classification accuracy of probit, ID3, and neural networks* , 1992 .

[8]  R. McKelvey,et al.  A statistical model for the analysis of ordinal level dependent variables , 1975 .

[9]  Donald F. Beal.  Advances in computer chess , 1986 .

[10]  Donald Michie,et al.  Expert systems in the micro-electronic age , 1979 .

[11]  Kar Yan Tam,et al.  A Comparative Analysis of Inductive-Learning Algorithms , 1993 .

[12]  Kathleen M. Swigger,et al.  Analysis of the Effectiveness of Machine Learning in Determining Decision Rules for Executive Compensation Planning , 1994 .

[13]  Pei-Lei Tu,et al.  Toward an intelligent classification-tree approach to problem-solving , 1989 .

[14]  Ingoo Han,et al.  Integrating statistical and inductive learning methods for knowledge acquisition , 1990 .

[15]  J. D. F. Habbema,et al.  Comparative performance of four discriminant analysis methods for mixtures of continuous and discrete variables , 1983 .

[16]  J. Ross Quinlan,et al.  Learning Efficient Classification Procedures and Their Application to Chess End Games , 1983 .

[17]  G. Judge,et al.  The Theory and Practice of Econometrics , 1981 .

[18]  William R. Dillon,et al.  On the Performance of Some Multinomial Classification Rules , 1978 .

[19]  David Perry Greene.  Automated Knowledge Acquisition: Overcoming the Expert System Bottleneck , 1987, ICIS.

[20]  J. D. F. Habbema,et al.  A Simulation Study of the Performance of Five Discriminant Analysis Methods for Mixtures of Continuous and Binary Variables , 1985 .

[21]  W. Hays.  Statistics for the social sciences , 1973 .

[22]  F. H. C. Marriott,et al.  Discrimination with Mixed Binary and Continuous Data , 1982 .

[23]  John R. Anderson,et al.  MACHINE LEARNING An Artificial Intelligence Approach , 2009 .

[24]  E. Feigenbaum,et al.  Applications of artificial intelligence for chemical inference. 22. Automatic rule formation in mass spectrometry by means of the meta-DENDRAL program , 1976 .

[25]  John Mingers,et al.  Expert Systems—Rule Induction with Statistical Data , 1987 .

[26]  James V. Hansen,et al.  Inducing rules for expert system development: an example using default and bankruptcy data , 1988 .

[27]  Joshua Lederberg,et al.  APPLICATIONS OF ARTIFICIAL INTELLIGENCE FOR CHEMICAL INFERENCE. 22. AUTOMATIC C RULE FORMATION IN MASS SPECTROMETRY BY MEANS OF THE META-DENDRAL PROGRAM , 1976 .

[28]  John S. Chandler,et al.  PREDICTING STOCK MARKET BEHAVIOR THROUGH RULE INDUCTION: AN APPLICATION OF THE LEARNING‐FROM‐EXAMPLE APPROACH* , 1987 .

[29]  T. Niblett,et al.  AUTOMATIC INDUCTION OF CLASSIFICATION RULES FOR A CHESS ENDGAME , 1982 .