Hampshire College

About: Hampshire College is a education organization based out in Amherst Center, Massachusetts, United States. It is known for research contribution in the topics: Genetic programming & Population. The organization has 461 authors who have published 998 publications receiving 40827 citations.

Effects of Lexicase and Tournament Selection on Diversity Recovery and Maintenance

Abstract: In genetic programming systems, parent selection algorithms select the programs from which offspring will be produced by random variation and recombination. While most parent selection algorithms select programs on the basis of aggregate performance on multiple test cases, the lexicase selection algorithm considers each test case individually, in random order, for each parent selection event. Prior work has shown that lexicase selection can produce both more diverse populations and more solutions when applied to several hard problems. Here we examine the effects of lexicase selection, compared to those of the more traditional tournament selection algorithm, on population error diversity using two program synthesis problems. We conduct experiments in which the same initial population is used to start multiple runs, each using a different random number seed. The initial populations are extracted from genetic programming runs, and fall into three categories: high diversity populations, low diversity populations, and populations that occur after diversity crashes. The reported data shows that lexicase selection can maintain high error diversity and also that it can re-diversify less-diverse populations, while tournament selection consistently produces lower diversity.

Assessing display variability in wild brown anoles Anolis sagrei using a mechanical lizard model

Abstract: Signals used for species identity ought to be highly stereotyped so as to facilitate immediate recognition by con- specifics. It is surprising therefore to find variability in putative species signature displays. The brown anole Anolis sagrei has a high degree of variability in its signature bobbing display. In this study we collected descriptive data on variability in the temporal structure of wild brown anole bobbing patterns, finding that no two displays analyzed had the same temporal structure, and we also tested whether wild brown anoles prefer the signature display over an alternate display pattern by using mechanical robot playbacks in the field. As a response metric we assessed whether or not the lizards showed social responses (pushup, dewlap ex- tension, or head-nod) in response to the robotic presentations. We found that the lizards responded slightly more to the signature than to the alternate pattern, providing support for the idea that despite the variability seen in displays, the signature pattern is meaningful to them. We tested two other independent variables: speed of the display and elevation of the robot during its display, neither of which was significant. Dewlap extensions were given predominantly by adult males and were more likely to be given in the breeding season than the nonbreeding season. Pushups and head-nods were given equally by males and a combined class of females and juveniles, and were not seasonal. Head-nods increased after the robot turned off, suggesting that they may be used in a conversational turn-taking style during communication (Current Zoology 57 (2): 140-152, 2011).