undefined

Deep Belief Network with Fuzzy Parameters and Its Membership Function Sensitivity Analysis

Publiceringsår

2025

Upphovspersoner

Shukla, Amit K.; Muhuri, Pranab K.

Abstrakt

Over the last few years, deep belief networks (DBNs) have been extensively utilized for efficient and reliable performance in several complex systems. One critical factor contributing to the enhanced learning of the DBN layers is the handling of network parameters, such as weights and biases. The efficient training of these parameters significantly influences the overall enhanced performance of the DBN. However, the initialization of these parameters is often random, and the data samples are normally corrupted by unwanted noise. This causes the uncertainty to arise among weights and biases of the DBNs, which ultimately hinders the performance of the network. To address this challenge, we propose a novel DBN model with weights and biases represented using fuzzy sets. The approach systematically handles inherent uncertainties in parameters resulting in a more robust and reliable training process. We show the working of the proposed algorithm considering four widely used benchmark datasets such as: MNSIT, n-MNIST (MNIST with additive white Gaussian noise (AWGN) and MNIST with motion blur) and CIFAR-10. The experimental results show superiority of the proposed approach as compared to classical DBN in terms of robustness and enhanced performance. Moreover, it has the capability to produce equivalent results with a smaller number of nodes in the hidden layer; thus, reducing the computational complexity of the network architecture. Additionally, we also study the sensitivity analysis for stability and consistency by considering different membership functions to model the uncertain weights and biases. Further, we establish the statistical significance of the obtained results by conducting both one-way and Kruskal-Wallis analyses of variance tests.
Visa mer

Organisationer och upphovspersoner

Vasa universitet

Shukla Amit Kumar

Publikationstyp

Publikationsform

Artikel

Moderpublikationens typ

Tidning

Artikelstyp

En originalartikel

Målgrupp

Vetenskaplig

Kollegialt utvärderad

Kollegialt utvärderad

UKM:s publikationstyp

A1 Originalartikel i en vetenskaplig tidskrift

Publikationskanalens uppgifter

Förläggare

Elsevier

Volym

614

Artikelnummer

128716

Publikationsforum

63879

Publikationsforumsnivå

2

Öppen tillgång

Öppen tillgänglighet i förläggarens tjänst

Ja

Öppen tillgång till publikationskanalen

Delvis öppen publikationskanal

Parallellsparad

Ja

Övriga uppgifter

Vetenskapsområden

Statistik; Data- och informationsvetenskap

Nyckelord

[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object],[object Object]

Publiceringsland

Nederländerna

Förlagets internationalitet

Internationell

Språk

engelska

Internationell sampublikation

Ja

Sampublikation med ett företag

Nej

DOI

10.1016/j.neucom.2024.128716

Publikationen ingår i undervisnings- och kulturministeriets datainsamling

Ja