SIMULATION OF SPATIAL-LEARNING IN THE MORRIS WATER MAZE BY A NEURAL-NETWORK MODEL OF THE HIPPOCAMPAL-FORMATION AND NUCLEUS-ACCUMBENS

Authors
Citation
Ma. Brown et Pe. Sharp, SIMULATION OF SPATIAL-LEARNING IN THE MORRIS WATER MAZE BY A NEURAL-NETWORK MODEL OF THE HIPPOCAMPAL-FORMATION AND NUCLEUS-ACCUMBENS, Hippocampus, 5(3), 1995, pp. 171-188
Citations number
75
Categorie Soggetti
Neurosciences
Journal title
ISSN journal
10509631
Volume
5
Issue
3
Year of publication
1995
Pages
171 - 188
Database
ISI
SICI code
1050-9631(1995)5:3<171:SOSITM>2.0.ZU;2-Y
Abstract
Cells in the hippocampal formation show spatial firing correlates thou ght to be critical to the role played by this structure in spatial lea rning. Place cells in the hippocampus proper show location-specific ac tivity, whereas cells in the postsubiculum fire as a function of momen tary directional heading. One question which has received little atten tion is how these spatial signals are used by motor structures to actu ally guide spatial behavior. Here we present a model of how one kind o f spatial behavior, instrumental learning in the Morris water maze, co uld be guided by the spatial information in the hippocampal formation. For this, we concentrate on the hippocampal projection to the nucleus accumbens, which is strongly implicated in instrumental learning. In the model, simulated firing patterns of place cells and head direction cells activate ''motor'' cells in the ''accumbens.'' Each motor cell causes a particular locomotor movement in a simulated rat. In this way , the ''rat'' locomotes through the simulated environment. Each step p laces the animal in a slightly different location and directional orie ntation, which, in turn, activates a different set of place and head d irection cells, thus causing the next locomotor response, and so on. C onnection strengths between cells are initially set randomly. When the animal encounters the reward location, however, connections are alter ed, so that recently active synapses are strengthened. Thus, successfu l moves in a particular locational and directional context are ''stamp ed in.'' Simulated rats show rapid learning, similar in many ways to t hat of actual rats. in particular, they generate efficient routes to t he goal after minimal experience, and can do so from somewhat novel st arting positions. Consideration of the model architecture shows that 1 ) combined use of directional and place information is an example of a linearly inseparable problem and that 2) some types of novel route ge neration, often thought to require a ''cognitive mapping'' strategy, c an be generated from the S-R type model used here. (C) 1995 Wiley-Liss , Inc.