C L U S P A C C O M P U T E R P R O G R A M S F O R C L U S T E R A N A L Y S I S --- probability-based, maximum likelihood clustering using the mixture-of-populations model --- common or varying covariance-matrix options --- adaptive metric clustering --- routines based on the "classification" likelihood and the "mixture" likehood --- includes an improved version of ISODATA --- source programs in FORTRAN ..................................................................... CLUSPAC Dr. Stanley L. Sclove IDS Dept. (MC 294) University of Illinois at Chicago Chicago, IL 60607-7124 (312) 996-2676 fax (312) 413-0385 ..................................................................... CLUSPAC Copyright (C) 1991 Stanley Louis Sclove ..................................................................... C L U S P A C: Sample Output --------------------------------------------------------------------- ..................................................................... MIXPDTA CLUSPAC - MIXTURE-MODEL CLUSTERING OF CASES VARYING COVARIANCE MATRICES AUTOMATIC SETTING OF INITIAL PARAMETER ESTIMATES Copyright (C) 1991 Stanley L. Sclove Developed and programmed by: Prof. Stanley L. Sclove, Ph.D. (312) 996-2681 Information & Decision Sciences Dept. M/C 294 College of Business Administration University of Illinois at Chicago Box 4348, Chicago, IL 60680-4348 Program Version: 2.8 01-Jan-90 (VM/CMS) ..................................................................... PROBLEM TITLE IS WBC DATA: COMP STUDY OF MACHINE VS MANUAL (WBC DATSORT) 6/ 5/1990 AT 16: 4:33 NUMBER OF VARIABLES . . . . . . . . . . . . 2 NUMBER OF OBSERVATIONS (SAMPLE SIZE), N . . 62 MINIMUM FOR EACH VARIABLE: 0.1 0.010 MAXIMUM FOR EACH VARIABLE: 83.1 81.500 @UNIT FIRST FOUR DATA POINTS ----- ---- ---- ------ 0.1 0.025 CS01 0.2 0.020 CS02 0.3 0.050 CS03 0.4 0.010 CS04 @END NUMBER OF CLUSTERS TO REPORT. . . . . . . . 1 TO 9 STATISTICS FOR WHOLE (UNCLUSTERED) SAMPLE: MEAN VECTOR: 7.17 7.08 COVARIANCE MATRIX: 119.697 116.815 116.815 114.726 MINUS 2 LOG LIKELIHOOD = 628.5 NUMBER OF PARAMETERS = 5 AIC = 638.5 SCHWARZ CRITERION = 649.1 ..................................................................... K = 2 CLUSTERS NUMBER OF PARAMETERS = 11 INITIAL MEANS MEAN VECTOR FOR CLUSTER 1: 0.1 0.025 MEAN VECTOR FOR CLUSTER 2: 5.5 5.400 ITERATION 1 MINUS 2 LOG LIKELIHOOD = 910.4 AIC = 932.4 SCHWARZ CRITERION = 955.8 CASE POSTERIOR PROBS OF CLUSTER MEMBERSHIP: @UNIT POSTERIOR PROB OF GROUP MEMBERSHIP FOR 1ST 4 CASES: 0.54 0.46 0.54 0.46 0.54 0.46 0.54 0.46 @END NUMBERS IN CLUSTERS: 19 43 . . . ITERATION 5 MINUS 2 LOG LIKELIHOOD = 540.4 AIC = 562.4 SCHWARZ CRITERION = 585.8 CASE POSTERIOR PROBS OF CLUSTER MEMBERSHIP: POSTERIOR PROB OF GROUP MEMBERSHIP FOR 1ST 4 CASES: 0.97 0.03 0.97 0.03 0.97 0.03 0.97 0.03 MIXING PROBABILITIES: 0.80 0.20 MEAN VECTOR FOR CLUSTER 1: 4.433 4.480 MEAN VECTOR FOR CLUSTER 2: 18.048 17.414 COMMON COVARIANCE MATRIX (MLE): 24.054 23.932 23.932 24.099 CONVERGENCE: NO CASE CHANGED CLUSTERS AFTER ITERATION 5. CLUSTER 1 CS01 CS02 CS03 CS04 CS05 CS06 CS07 CS08 CS09 CS10 CS11 CS12 CS13 CS14 CS15 CS16 CS17 CS18 CS19 CS20 CS21 CS22 CS23 CS24 CS25 CS26 CS27 CS28 CS29 CS30 CS31 CS32 CS33 CS34 CS35 CS36 CS37 CS39 CS40 CS41 CS42 CS43 CS44 CS45 CS46 CS48 CS49 CS50 CS51 CS52 CS53 CLUSTER 2 CS38 CS47 CS54 CS55 CS56 CS57 CS58 CS59 CS60 CS61 CS62 ................................................................................ K = 3 CLUSTERS NUMBER OF PARAMETERS = 17 INITIAL MEANS MEAN VECTOR FOR CLUSTER 1: 0.1 0.025 MEAN VECTOR FOR CLUSTER 2: 2.7 3.500 MEAN VECTOR FOR CLUSTER 3: 6.8 7.100 ITERATION 1 MINUS 2 LOG LIKELIHOOD = 1056.4 AIC = 1090.4 SCHWARZ CRITERION = 1126.5 @UNIT CASE POSTERIOR PROBS OF CLUSTER MEMBERSHIP: POSTERIOR PROB OF GROUP MEMBERSHIP FOR 1ST 4 CASES: 0.40 0.51 0.10 0.40 0.51 0.10 0.39 0.51 0.10 0.39 0.51 0.10 @end NUMBERS IN CLUSTERS: 0 42 20 . . . ................................................................................ @UNIT WBC DATA: COMP STUDY OF MACHINE VS MANUAL (WBC DATSORT) MODEL SELECTION CRITERIA MODEL-SELECTION CRITERIA NUMBER MINUS 2 NUMBER (PPH=POST.PROB.OF MODEL): OF TIMES OF AKAIKE SCHWARZ CLUSTERS, LOG OF PARAM- ------------ ------------ K LIKELIHOOD ETERS VALUE PPH VALUE PPH --------- ---------- ----- ------- --- ------- --- 1 628.5 5 638.5 0.00 649.1 0.00 2 540.4 11 562.4 0.00 585.8 0.00 3 490.0 17 524.0 0.00 560.2 0.00 4 475.3 23 521.3 0.00 570.3 0.00 5 408.0 29 466.0 1.00 527.7 1.00 6 457.0 35 527.0 0.00 601.4 0.00 7 484.6 41 566.6 0.00 653.8 0.00 8 430.0 47 524.0 0.00 624.0 0.00 9 450.7 53 556.7 0.00 669.4 0.00 ................................................................................ @end WBC DATA: COMP STUDY OF MACHINE VS MANUAL (WBC DATSORT) CONDITION CODE FOR THIS RUN: PROGRAM ENDED NORMALLY. ................................................................................ MIXPCMA CLUSPAC - MIXTURE-MODEL CLUSTERING OF CASES TIME BEGAN: 6/ 5/1990 AT 16: 4:33 TIME FINISHED: 6/ 5/1990 AT 16: 4:35