Design and Implementation of Neural Networks Neurons with RadBas, LogSig, and TanSig Activation Functions on FPGA

摘要

Artificial Neural Networks (ANNs) are the computational models which copy the human brain's learning and decision making abilities. Before they are used, they are gone through a learning process and later are utilized in decision making processes. Their decision making ability and correctness of the decisions depend on several key factors such as the structure of network, type of activation functions in the neurons, period of learning, and number of stimuluses used during the learning process [1].