The Center for Biomedical Informatics
State University of Campinas, Brazil

Research Abstracts

The identification, isolation and quantification of animal behavioral patterns and sequences represent essential steps for the analysis of observed behavior using the ethological approach. Several statistical techniques have been developed for this purpose, based on Markov chains, point processes, grammatical models, cluster analysis, etc. Many of the difficulties and poor performance associated to these techniques can be traced to time-varying transition probabilities between elements of a sequence and the essential non-linear separability of multidimensional patterns. In this paper we present a novel approach to the problem of pattern classification using artificial neural networks. Artificial neural networks, or connectionist systems, are being increasingly used to represent and to process information by means of networks of interconnected processing elements, similar to neurones. Several emerging global properties of connectionist systems, such as associative memory, distributed parallel processing, learning, etc.; have favored its applications in a large variety of tasks involving pattern classification and recognition. Neural networks are easier to implement and to train using examples, rather than complex and unreliable heuristics. They do not require rigid analytical or causal assumptions, such as parametric distributions or linear separability of patterns, have a higher reliability in presence of noise and uncertainty and are able to provide graded or fuzzy responses. Complex nonlinear as well as time-ordered phenomena are easily treated by ANNs, in contrast to other approaches. In addition, ANNs have the strong incentive that, once true massively-parallel neurocomputers are available, they will provide unprecedented, blinding-fast devices to implement intelligent applications in this field. We have developed and tested a multilayer perceptron (MLP) neural network architecture, which is capable of learning behavioral pattern recognition tasks using the error backpropagation algorithm. A computer program for IBM-PC compatible microcomputers, named NEURONET was developed in order to implement a sequential simulation of the MLP. A database consisting of previously classified behavioral patterns derived from the observation of aggressive, defensive and escape behavior of brain-stimulated cats, was submitted to the MLP neural network for training; using the following procedure. Each input node of the MLP was assigned to a different behavioral item in the predefined cat ethogram. A total of 20 behavioral, motor and autonomic reaction items were identified, and the presence/absence of a behavioral item in the input pattern were defined as 1 or 0 numerical inputs to the corresponding neurone, respectively. Each output node of the MLP was assigned to a pattern classification. A total of 3 behavioral patterns were identified, and the correct output pattern for a given input pattern was identified by the observer by setting the value 1 for its corresponding output node, while all the others were set to zero. The training set was presented several times to the neural network, until learning was achieved (defined as a mean classification error of less than 2 % of the training patterns). Finally, a second set of previously classified patterns was tested with the trained network, and its overall classification accuracy evaluated. The neural network was able to classify these patterns with a 90 % accuracy. Interesting enough, mixed behavioral patterns (e.g., aggression/defense) were thus identified by the neural network, as evidenced by the fact that two or more output nodes in the network had significant values above zero. The present work provides a demonstration of the usefulness and viability of the ANN approach in computational analysis of observed behavioral patterns in ethology.

Published int:

Return to Home Page

Return to Abstracts Index

Last Updated: March 2, 1996

renato@sabbatini.com