Subgrouped Real Time Recurrent Learning Neural Networks
Abstract
A subgrouped Real Time Recurrent Learning (RTRL) network was evaluated. The one layer net successfully learns the XOR problem, and can be trained to perform time dependent functions. The net was tested as a predictor on the behavior of a signal, based on past behavior. While the net was not able to predict the signal's future behavior, it tracked the signal closely. The net was also tested as a classifier for time varying phenomena; for the differentiation of five classes of vehicle images based on features extracted from the visual information. The net achieved a 99.2% accuracy in recognizing the five vehicle classes. The behavior of the subgrouped RTRL net was compared to the RTRL network described in Capt R. Lindsey's AFIT Master's thesis. The subgrouped RTRL performance proved close to the RTRL network in accuracy while reducing the time required to train the network for multiple output (classification) problems. Neural network, Recurrent, RTRL, Image recognition time dependence, Temporal, Sequence recognition.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1994
- Accession Number
- ADA280618
Entities
People
- Jeffrey S. Dean
Organizations
- Air Force Institute of Technology