J. Starzyk and Z. Zhu (USA)
Self-Organizing Learning Array, Software Simulation, Self-Organizing Hardware/Fabric
A neural network paradigm named Self-Organizing Learning Array system has been simulated in software. Hardware implementation limitations are considered in this simulation. Simulation is performed based on a benchmark classification problem, the Australian Credit Card problem. The system behavior is observed and the learning algorithm is examined. The correct classification rate has been compared with some existing classification methods. Although not particularly designed for solving this type of classification problem, this system still shows very good performance. The system will be implemented in FPGA and SOC.
Important Links:
Go Back
