Home About Login Current Archives Announcements Editorial Board
Submit Now For Authors Call for Submissions Statistics Contact
Home > Archives > Volume 20, No 8 (2022) > Article

DOI: 10.14704/nq.2022.20.8.NQ44729

GENERALIZED FUNCTIONAL ADDITIVE MODEL FOR VALIDATING THE SPIKE TRANSFORMATION OF FNDCNN USING 2D IMAGES

Dr.P.Nancy, T.M.Angelin Ben Roja, Dr.E.A.Mohamed Ali

Abstract

Understanding, enhancing, healing, and repairing the neural system is a predominant and fastgrowing research area of neural engineering and neuroscience. The Convolution Neural Network (CNN/ConvNet) is a popular and preferred choice of researchers to identify the neural functions because of its attractive features in image recognition. Moreover, the linear nature of CNN reduces its effectiveness to identify the neural activity based on image classification. Hence, a novel Fuzzy Logic Non-linear Deep Convolution Neural Network (FNDCNN) is proposed in this paper to overcome the limitation of conventional CNN. The proposed novel deep learning FNDCNN approach is used to model neural spiking activity of the brain cells that is able to predict the output neural spiking activity from the input. The nonlinear activation dynamic of the FNDCNN is introduced by fuzzy logic function in higher-order kernels. The performance of the proposed novel deep learning approach designed for Multi-input Multi-output (MIMO) system is tested and compared with the recent techniques such as conventional CNN, MicroNet, and GLM in terms of correlation coefficients and Normalized Root Mean Square Error (NRMSE) between actual and predicted output neural spiking activity. The proposed FNDCNN algorithm improves the accuracy and performance of the MIMO system model and also ensures better results when compared with the conventional CNN, Micronet, and GFM

Keywords

Non Linear CNN, Neural spiking, Fuzzy, Generalized functional additive model

Full Text

PDF

References

?>