Abstract. This paper investigates the effects of early stopping as a method to counteract overfitting in evolutionary data modelling using Genetic Programming. Early stopping has been proposed as a method to avoid model overtraining, which has been shown to lead to a significant degradation of out-of-sample performance. If we assume some sort of performance metric maximisation, the most widely used early training stopping criterion is the moment within the learning process that an unbiased estimate of the performance of the model begins to decrease after a strictly monotonic increase through the earlier learning iterations. We are conducting an initial investigation on the effects of early stopping in the performance of Genetic Programming in symbolic regression and financial modelling. Empirical results suggest that early stopping using the above criterion increases the extrapolation abilities of symbolic regression models, but is by no means the optimal training-stopping criterion in the case of a real-world financial dataset.

No abstract

No abstract

Following a recent call for a suite of benchmarks for genetic programming [4], we investigate the criteria for a meaningful dynamic benchmark for GP. We explore the design of a dynamic benchmark for symbolic regression, based on semantic distance between evaluated functions, where larger distances serve as a proxy for greater environmental change. We do not find convincing evidence that lower semantic distance is a good proxy for greater ease in adapting to a change. We conclude that due to fundamental characteristics of GP, it is difficult to come up with a single dynamic benchmark problem which is generally applicable.