Lexicase selection is a recent parent selection technique originally developed by Lee Spector and Thomas Helmuth for solving program synthesis problems. It promotes solutions that perform well on uncommon sets of training cases. I developed epsilon-Lexicase selection, which extends this algorithm to continuous data. It has demonstrated strong results on regression problems [1]. Recently I have focused on developing the theory underlying lexicase and epsilon-lexicase selection using probabilistic and multi-objective approaches [2].

  1. La Cava, W., Spector, L., & Danai, K. (2016, July). Epsilon-lexicase selection for regression. In Proceedings of the 2016 on Genetic and Evolutionary Computation Conference (pp. 741-748). ACM. preprint

  2. La Cava, W., Helmuth, T., Spector, L., & Moore, J. H. (2017). epsilon-Lexicase selection: a probabilistic and multi-objective analysis of lexicase selection in continuous domains. arXiv