Differential Evolution for Rule Extraction and Its application in Recognizing Oil Reservoir

The method of rule extraction based on differential evolution (DE-Rule) is proposed in this paper. The style of rule is IF-THEN. Connection word is AND in the antecedent of rule. The consequence of rule is class label. The above rule is encoded and expressed to individual in population of differential evolution. Then the population is iterated through differential mutation and binomial crossover. Optimal rule set is obtained after decoding the optimal individual. Finally, DE-Rule, RS-Rule and ANN-GA-Cascades-Rule are used to recognize oil reservoir (dry layer, water layer, inferiority layer and oil layer) in the Jianghan oil field. Data of well oilsk81 is training data and data of well oilsk83 is testing data. The results show that performance of accuracy and interpretability in DE-Rule is the best.