Bio-Inspired Strategies for Modeling and Detection in Diabetes Mellitus Treatment

3: Mathematical preliminaries

Abstract

This chapter introduces the bio-inspired tools used in this book, such as evolutionary algorithms, particle swarm-based algorithms, neural networks, and deep neural networks. Regarding evolutionary algorithms, Evonorm is shown, which replaces the crossover and mutation stages by a distribution function in addition to the differential evolution algorithm. In the algorithms based on swarms of particles, the optimization algorithm by swarm of particles and the optimization algorithm by a colony of ants are shown. The multilayer Perceptron neural networks and the discrete high-order neural network are shown. Deep neural networks include the convolutional neural network, LSTM, BiLSTM, LSTM-FCN, and ResNet.

Keywords ...