A novel recursive gene selection method based on least square kernel extreme learning machine
Abstract : This paper presents a recursive feature elimination (RFE) mechanism to select the most informative genes with a least square kernel extreme learning machine (LSKELM) classifier.Describing the generalization ability of LSKELM in a way that is related to small norm of weights, we proposed a ranking criterion to evaluate the importance of genes by the norm of weights obtained by LSKELM network.The proposed method is called LSKELM-RFE algorithm,which first employs the original genes to build a LSKELM classifier, and then ranks the genes according to their importance given by the norm of LSKELM network output weights, and finally removes a least important gene.Benefiting from the random mapping mechanism of the extreme learning machine (ELM) kernel, there are no parameter of LSKELM-RFE needs to be manually tuned.A comparative study among our proposed algorithm and other two famous RFE algorithms has shown that LSKELM-RFE outperforms other RFE algorithms in both the computational cost and generalization ability.
• All existing one-class classifiers, which use activation function, that can work only for differentiable function but our all proposed methods in the expanded toolbox can work for both differentiable and non-differentiable activation function.
• We propose an expansion of DD toolbox for both online and offline OCC based on ELM and OSELM.
• As we have discussed, existing autoencoder is a time consumable method during training because it needs to tune the weight properly but our all three proposed autoencoders based methods have no need to tune the weight and yield the result in just one pass
These nonlinear entities are usually evolved into the sophisticated computer programs to solve a particular problem. Many researches show that GEP has very strong ability in data mining and optimization, and is especially suitable for dealing with function mining and symbolic regression problems.
? GEP chromosome is represented by a fixed-length character string containing one or more genes. Each gene contains two parts: head and tail. The head can be composed of functions and terminals, while the tail can only consist of terminals.
? The function set is formed by all function symbols needed for solving the object problem while the terminals set consist of known symbols, variable and constant described the problem.
• Our proposed classifiers exhibit better performance compared to ten traditional one-class classifiers and ELM based two one-class classifiers.
• Through proposed one-class classifiers, we intend to expand the functionality of the most used toolbox for OCC i.e. DD toolbox.
• All of our methods are totally compatible with all the present features of the toolbox.
• We have expanded the toolbox of OCC provided by Tax, so we divided our proposed work based on the category provided by Tax only.
? The forecast performance is difficult to meet the needs of people simply using soft computing-based methods due to the difficulties in describing the large-scale weather process and useless hydrological information.
? Therefore, in consideration of the importance of rainfall in our daily life and the difficulties to predict it, this work was to develop a precipitation prediction model based on ELM and Gene Expression Programming (GEP) to improve regional daily precipitation prediction.
? The rest of the paper is organized as follows. Firstly, the related method theories are briefly introduced. Next, we clearly describes the proposed hybrid method
|