Research Of The Hidden Layer Structure Dynamic Adjustment Method Based On Convex Incremental Extreme Learning Machine

Extreme Learning Machine(ELM)is a generalized learning algorithm based on single-hidden-layer feed-forward neural network.Since ELM has an excellent performance on regression and classification problems,it has been paid more and more attention at home and abroad recently.Well we all know,the determination of structure of ELM plays a vital role in the applications.Essentially,determination of the structure of ELM is equivalent to the determination of the hidden layer structure.Utilizing a smaller scale of the hidden layer structure can promote the running speed greatly.With a thorough analysis on the current classical related adjustable ELMs,in this paper,we propose algorithm PCI-ELM(Pruned-Convex Incremental Extreme Learning Machine)based on CI-ELM(Convex Incremental Extreme Learning Machine).Furthermore,we also present an improved PCI-ELM algorithm,EPCI-ELM(Enhanced Pruned-Convex Incremental Extreme Learning Machine),which introduces a filtering strategy for PCI-ELM during the neural adding process.In order to adjust the single hidden layer feed-forward neural network more flexibly and achieve the most compact form of the hidden layer structure,in this paper,we get rid of the adjusting process depending on only one neural network,but propose a more dynamic hidden layer structure determining algorithm:DCI-ELM(Dynamic Convex Incremental Extreme Learning Machine).DCI-ELM adequately compares all current known hidden layer neurons to organize the most compact structure of the hidden layer.By using the currently known hidden layer neurons,we select the best hidden layer structure under each scale in the building process.In the same size,based on the competitive filtering,only the structure which has a better neural network learning accuracy w be reserved.In addition,each neural network on each scale is compared with the training goals when a new neuron is introduced.Therefore,the simplest neural network hidden layer structure will be chosen as the best one.At the end of this paper,we verify the performance of PCI-ELM,EPCI-ELM and DCI-ELM.The results show that PCI-ELM,EPCI-ELM and DCI-ELM control hidden layer structure very well and construct the more compact single-hidden-layer feed-forward neural network.