题目: Dendritic Neuron Models, Learning Algorithms and Applications

摘要: An artificial neural network (ANN) that mimics the information processing mechanisms and procedures of neurons in human brains has achieved great success in many fields, e.g., classification, prediction and control. However, traditional ANNs suffer from many problems, such as the hard understanding problem, the slow and difficult training problem and the difficulty to scale them up. These drawbacks motivate us to develop a new dendritic neuron model (DNM) by considering the nonlinearity of synapses, not only for a better understanding of a biological neuronal system, but also for providing a more useful method for solving practical problems. To achieve its better performance for solving problems, six learning algorithms including biogeography-based optimization, particle swarm optimization, genetic algorithm, ant colony optimization, evolutionary strategy and population-based incremental learning are used to train it instead of the traditional backpropagation algorithm. The best combination of its user-defined parameters has been systemically investigated by using the Taguchi’s experimental design method. The experiments on fourteen different problems involving classification, approximation and prediction are conducted by various neural networks. The results answer which one is the most effective in training them and prove the outstanding performance of DNM over other neural networks in solving classification, approximation and prediction problems. This talk will also reveal the novel combination of DNM and a decision-tree-based initialization method and its application to semiconductor manufacturing equipment’s fault diagnosis

报告人: MengChu Zhou，Distinguished Professor in Electrical and Computer Engineering, New Jersey Institute of Technology (NJIT), Newark, NJ.

时间：9月20日（周日）下午9:00-12:00

地点：腾讯会议，ID：794 686 742

欢迎师生参加！