Abstract
Neural networks have always been a popular approach for intelligent machine development and knowledge discovery. Although, reports have featured successful neural network implementations, problems still exists with this approach, particularly its excessive training time. In this paper, we propose a Gene-Regulated Nested Neural Network (GRNNN) model as an improvement to existing neural network models to solve the excessive training time problem. We use a Gene Regulatory Training Engine (GRTE) to control and distribute the genes that regulate the proposed nested neural network. The proposed GRNNN is evaluated and validated through experiments to classify accurately the 8bit XOR parity problem. Experimental results show that the proposed model does not require excessive training time and meets the required objectives.