Preliminary to a Neural Network Model of Sonar-Based Target Discrimination in the Echolocating Bat
Abstract
It's known from previous research (Griffin, 1958) that echolocating bats such as the big brown bat Eptesicus fuscus, can discriminate between edible and inedible airborne targets using information carried in the echo returns. The goal of this project is to build a neural network model which can perform a rudimentary discrimination task using the same sonar targets. The model is to serve as a preliminary test of the network's ability to discriminate, categorize, and generalize from a limited database. We plan to use this model in a later study comparing the performance of our network with the behavioral data collected by Griffin, trying to duplicate as closely as possible the parameters of the task originally presented to bats.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1988
- Accession Number
- ADA205681
Entities
People
- Cameron Brennan
- Lynda Chen
Organizations
- Brown University