Stability and task complexity: a neural network model of genetic assimilation

Since Hinton and Nowlan introduced the Baldwin effect to the evolutionary computation community, agent-based studies of genetic assimilation have uncovered many details of the dynamic processes involved. In a previous paper, we demonstrated genetic assimilation with a simple food/toxin discrimination task using neural network agents that could evolve their learning rate. The study reported in this paper investigated the genetic assimilation of more complex learning tasks. Kauffman's NK landscape model, which can generate landscapes with a variable degree of correlation, was used to define learning tasks of varying levels of complexity. Simulations indicate an increased tendency of genetic assimilation to occur as the complexity of the learning task decreases and the environmental stability increases. These results are explained in terms of the shifting balance between the evolutionary costs and benefits of learning.