IEEE Computational Intelligence Magazine - February 2023 - 17

Yongduan Song
Chongqing University, CHINA
Publication
Spotlight
Dongrui Wu
Huazhong University ofScience and Technology, CHINA
Carlos A. Coello Coello
CINVESTAV-IPN, MEXICO
Georgios N. Yannakakis
University ofMalta, MALTA
Huajin Tang
Zhejiang University, CHINA
Yiu-ming Cheung
Hong Kong Baptist University, HONG KONG
Hussein Abbass
University ofNew South Wales, AUSTRALIA
CIS Publication Spotlight
IEEE Transactions on Neural
Networks andLearning Systems
AGradient-Guided Evolutionary Approach
to Training Deep Neural Networks,
by S. Yang, Y.Tian,C.He,X.
Zhang, K. C. Tan, and Y. Jin,
IEEE Transactions on Neural Networks
and Learning Systems,Vol.33,
No. 9, Sep. 2022, pp. 4861-4875.
Digital Object Identifier: 10.1109/
TNNLS.2021.3061630
" It has been widely recognized that
the efficient training of neural networks
(NNs) is crucial to classification performance.
While a series of gradient-based
approaches have been extensively
developed, they are criticized for the
ease of trapping into local optima
and sensitivity to hyperparameters. Due
to the high robustness and wide applicability,
evolutionary algorithms (EAs)
have been regarded as a promising alternative
for training NNs in recent years.
However, EAs suffer from the curse of
Digital Object Identifier 10.1109/MCI.2022.3223500
Date ofcurrent version: 13January 2023
training DNNs but also enhances the performance
of EAs in solving large-scale
optimization problems. "
IEEE Transactions on Fuzzy Systems
IMAGE LICENSED BY INGRAM PUBLISHING
dimensionality and are inefficient in
training deep NNs (DNNs). By inheriting
the advantages ofboth the gradientbased
approaches and EAs, this article
proposes a gradient-guided evolutionary
approach to train DNNs. The proposed
approach suggests a novel genetic operator
to optimize theweights in the search space,
where the search direction is determined
by the gradient ofweights. Moreover, the
network sparsity is considered in the proposed
approach, which highly reduces
the network complexity and alleviates
overfitting. Experimental results on singlelayer
NNs, deep-layer NNs, recurrent
NNs, and convolutional NNs (CNNs)
demonstrate the effectiveness of the proposed
approach. In short, this work not
only introduces a novel approach for
Multiclass Fuzzily Weighted AdaptiveBoosting-Based
Self-Organizing Fuzzy
Inference Ensemble Systems for Classification,
by X. Gu and P. P. Angelov,
IEEE Transactions on Fuzzy Systems,
Vol. 30, No. 9, Sep. 2022,
pp. 3722-3735.
Digital Object Identifier: 10.1109/
TFUZZ.2021.3126116
" Adaptive boosting (AdaBoost) is a
widely used technique to construct a
stronger ensemble classifier by combining
a set of weaker ones. Zero-order
fuzzy inference systems (FISs) are very
powerful prototype-based predictive
models for classification, offering both
great prediction precision and high user
interpretability. However, the use of
zero-order FISs as base classifiers in
AdaBoost has not been explored yet.
To bridge the gap, in this article, a novel
multiclass fuzzily weighted AdaBoost
FEBRUARY 2023 | IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE 17
IEEE Computational Intelligence Magazine - February 2023

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - February 2023
