Symbolic regression driven by training data and prior knowledge

Abstract

In symbolic regression, the search for analytic models is typically driven purely by the prediction error observed on the training data samples. However, when the data samples do not sufficiently cover the input space, the prediction error does not provide sufficient guidance toward desired models. Standard symbolic regression techniques then yield models that are partially incorrect, for instance, in terms of their steady-state characteristics or local behavior. If these properties were considered already during the search process, more accurate and relevant models could be produced. We propose a multi-objective symbolic regression approach that is driven by both the training data and the prior knowledge of the properties the desired model should manifest. The properties given in the form of formal constraints are internally represented by a set of discrete data samples on which candidate models are exactly checked. The proposed approach was experimentally evaluated on three test problems with results clearly demonstrating its capability to evolve realistic models that fit the training data well while complying with the prior knowledge of the desired model characteristics at the same time. It outperforms standard symbolic regression by several orders of magnitude in terms of the mean squared deviation from a reference model. CCS CONCEPTS • Computing methodologies → Genetic algorithms; Model development and analysis; • Theory of computation → Design and analysis of algorithms; Genetic programming; • Applied computing → Multi-criterion optimization and decision-making;

Symbolic Regression Methods for Reinforcement Learning

“…Depending on the problem, the function set can be arbitrarily enriched, for example by adding trigonometric functions. Furthermore, it is easy to use additional prior knowledge and constraints in SR methods in order to generate models with desired properties, see [33], [34].…”

Shape-constrained Symbolic Regression with NSGA-III

“…Auguste et al presented two new methods to include monotonic constraints in regression and classification trees [1]. In [9] the authors present a multi-objective symbolic regression approach to minimize the approximation error on the training data as well as the constraint violations on the constraint dataset. Therefore, they extended the NSGA-II algorithm and used sampling to evaluate the constraints.…”

Using Shape Constraints for Improving Symbolic Regression Models

“…They use a satisfiability solver to check if each candidate model fulfills shape or symmetry constraints by looking for counter-examples of the desired constraint. Kubalík et al [19] use multi-objective GP to minimize the approximation error of the training data and minimize the constraints for constraint data. The constraint data are artificially generated to capture the desired constraint.…”