Improving Genetic Programming for Classification with Lazy Evaluation and Dynamic Weighting

Abstract

In the standard process of creating classification decision trees with genetic programming, the evaluation process it the most time-consuming part of the whole evolution loop. Here we introduce a lazy evaluation approach of classification decision trees in the evolution process, that does not evaluate the whole population but evaluates only the individuals that are chosen to participate in the tournament selection method. Further on, we used dynamic weights for the classification instances, that are linked to the chance of that instance getting picked for the evaluation process and are determined by that instance’s classification rate. These instance weights change based on the misclassification rate of the instance. We thoroughly describe and experiment with the lazy evaluation on standard classification benchmark datasets and show that not only lazy evaluation approach uses less time to evolve the good solution, but can even produce statistically better solution due to changing instance weights and thus preventing the overfitting of the solutions.