Premium
Modeling associative learning with generalization for a case of warning signals
Author(s) -
Higashi Shigeo Yachi and Masahiko
Publication year - 1999
Publication title -
ecological research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.628
H-Index - 68
eISSN - 1440-1703
pISSN - 0912-3814
DOI - 10.1046/j.1440-1703.1999.143299.x
Subject(s) - generalization , aposematism , associative learning , content addressable memory , associative property , computer science , intraspecific competition , artificial intelligence , population , predation , mathematics , cognitive psychology , predator , ecology , artificial neural network , psychology , biology , pure mathematics , mathematical analysis , demography , sociology
Animals’ associative learning plays a crucial role in many intraspecific or interspecific interactions, involving an animal’s use of information on its interacting counterparts. Here, we present a theoretical model that captures the basic features of an animal’s associative learning, which may involve generalization, for a simplest case of warning signals. Specifically, we derive formulae for the average level of associative memory as functions of a few parameters that reflect the population density of prey, predator’s efficiency of prey detection, and the properties of predator’s associative learning, including generalization and memory decay. This average level of associative memory is of central importance in determining prey’s fitness and, thus, the evolution of warning signals (i.e. aposematism). The derived formula also shows that another species with similar signal enhances the fitness of an aposematic species of concern as long as their signal is similar enough for generalization to occur. The model developed here can be extended to more complicated cases and the basic idea can be applied to modeling other interactions involving associative learning with generalization.