Latest changes

UCS/RESULTS_FILE.txt

- Accuracy: 0.78125 Total time: 114.03 Rules: 360
- Total Av. Rules: 247.06666666666666
- Accuracy: 0.78125 Total time: 114.62 Rules: 365
- Total Av. Rules: 488.4
+ Accuracy: 0.90625 Total time: 33.04 Rules: 199
+ Accuracy: 0.921875 Total time: 34.97 Rules: 194
+ Accuracy: 0.859375 Total time: 35.2 Rules: 197
+ Accuracy: 0.859375 Total time: 36.26 Rules: 193
+ Accuracy: 0.890625 Total time: 34.05 Rules: 205
+ Accuracy: 0.875 Total time: 34.18 Rules: 201
+ Accuracy: 0.921875 Total time: 34.91 Rules: 218
+ Accuracy: 0.90625 Total time: 35.54 Rules: 202
+ Accuracy: 0.890625 Total time: 35.22 Rules: 205
+ Accuracy: 0.828125 Total time: 37.87 Rules: 216

UCS/UCS_Configuration_File.txt

@@ -8,9 +8,9 @@ trainFile=6Multiplexer_Data_Complete.txt		# FileName of training dataset
 testFile=None									# FileName of testing dataset.  If no testing data available or desired, put 'None'.
 outputDirectory=Local_Output					# Output file directory
 outputFile=ExampleRun							# FileName of output files.
-learningIterations=20000					    # Specify complete algorithm evaluation checkpoints and maximum number of learning iterations (e.g. 1000.2000.5000 = A maximum of 5000 learning iterations with evaluations at 1000, 2000, and 5000 iterations)
-N=1000											# Maximum size of the rule population (a.k.a. Micro-classifier population size, where N is the sum of the classifier numerosities in the population)
-p_spec=0.7										# The probability of specifying an attribute when covering. (1-p_spec = the probability of adding '#' in ternary rule representations). Greater numbers of attributes in a dataset will require lower values of p_spec.
+learningIterations=10000					    # Specify complete algorithm evaluation checkpoints and maximum number of learning iterations (e.g. 1000.2000.5000 = A maximum of 5000 learning iterations with evaluations at 1000, 2000, and 5000 iterations)
+N=500										    # Maximum size of the rule population (a.k.a. Micro-classifier population size, where N is the sum of the classifier numerosities in the population)
+p_spec=0.6									# The probability of specifying an attribute when covering. (1-p_spec = the probability of adding '#' in ternary rule representations). Greater numbers of attributes in a dataset will require lower values of p_spec.
 kfold=5									        # if not used, set to 0.
 
 ######--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
@@ -26,9 +26,9 @@ trackingFrequency=0								# Specifies the number of iterations before each esti
 ######--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 ###### Supervised Learning Parameters - Generally just use default values.
 ######--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-nu=5											# (v) Power parameter used to determine the importance of high accuracy when calculating fitness. (typically set to 5, recommended setting of 1 in noisy data)
+nu=10											# (v) Power parameter used to determine the importance of high accuracy when calculating fitness. (typically set to 5, recommended setting of 1 in noisy data)
 chi=0.8											# (X) The probability of applying crossover in the GA. (typically set to 0.5-1.0)
-upsilon=0.4									    # (u) The probability of mutating an allele within an offspring.(typically set to 0.1-0.5)
+upsilon=0.5									    # (u) The probability of mutating an allele within an offspring.(typically set to 0.1-0.5)
 theta_GA=25										# The GA threshold; The GA is applied in a set when the average time since the last GA in the set is greater than theta_GA.
 theta_del=20									# The deletion experience threshold; The calculation of the deletion probability changes once this threshold is passed.
 theta_sub=20									# The subsumption experience threshold;

UCS/UCS_OutputFileManager.py

@@ -34,7 +34,6 @@ from UCS.UCS_Constants import *
 class OutputFileManager:
 
     totalPopulationSize = 0
-    totalAverage = 0
 
     def writePopStats(self, outFile, trainEval, testEval, exploreIter, pop, correct):
         """ Makes output text file which includes all of the evaluation statistics for a complete analysis of all training and testing data on the current LCS rule population. """
@@ -140,7 +139,6 @@ class OutputFileManager:
         # Write each classifier--------------------------------------------------------------------------------------------------------------------------------------
 
         OutputFileManager.totalPopulationSize = len(pop.popSet)
-        OutputFileManager.totalAverage += len(pop.popSet)
         print(len(pop.popSet))
         for cl in pop.popSet:
             rulePopOut.write(str(cl.printClassifier()))

UCS/UCS_Run.py

@@ -85,11 +85,10 @@ def mainRun():
     print("Total run time in seconds: %.2f" % total)
     f = open("RESULTS_FILE.txt", 'a')
     f.write(" Accuracy: " + str(kfold_accuracy) + " Total time: " + str(total) + " Rules: " + str(OutputFileManager.totalPopulationSize) + "\n")
-    f.write(" Total Av. Rules: " + str(OutputFileManager.totalAverage / 15) + "\n")
 
 if __name__ == '__main__':
 
-    for i in range(15):
+    for i in range(10):
         mainRun()