Deep Neural Fuzzy System Oriented toward High-Dimensional Data and Interpretable Artificial Intelligence

被引:7
作者
Chen, Dewang [1 ,2 ,3 ]
Cai, Jijie [1 ,3 ]
Huang, Yunhu [1 ,3 ]
Lv, Yisheng [4 ]
机构
[1] Fuzhou Univ, Coll Math & Comp Sci, Fuzhou 350108, Peoples R China
[2] Fujian Univ Technol, Sch Transportat, Fuzhou 350108, Peoples R China
[3] Fuzhou Univ, Key Lab Intelligent Metro Univ Fujian Prov, Fuzhou 350108, Peoples R China
[4] Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2021年 / 11卷 / 16期
基金
中国国家自然科学基金;
关键词
interpretability; high-dimensional data; adaptive neuro-fuzzy inference system; deep neural fuzzy system; MINIBATCH GRADIENT DESCENT; ANFIS; CLASSIFICATION; REGULARIZATION;
D O I
10.3390/app11167766
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Fuzzy systems (FSs) are popular and interpretable machine learning methods, represented by the adaptive neuro-fuzzy inference system (ANFIS). However, they have difficulty dealing with high-dimensional data due to the curse of dimensionality. To effectively handle high-dimensional data and ensure optimal performance, this paper presents a deep neural fuzzy system (DNFS) based on the subtractive clustering-based ANFIS (SC-ANFIS). Inspired by deep learning, the SC-ANFIS is proposed and adopted as a submodule to construct the DNFS in a bottom-up way. Through the ensemble learning and hierarchical learning of submodules, DNFS can not only achieve faster convergence, but also complete the computation in a reasonable time with high accuracy and interpretability. By adjusting the deep structure and the parameters of the DNFS, the performance can be improved further. This paper also performed a profound study of the structure and the combination of the submodule inputs for the DNFS. Experimental results on five regression datasets with various dimensionality demonstrated that the proposed DNFS can not only solve the curse of dimensionality, but also achieve higher accuracy, less complexity, and better interpretability than previous FSs. The superiority of the DNFS is also validated over other recent algorithms especially when the dimensionality of the data is higher. Furthermore, the DNFS built with five inputs for each submodule and two inputs shared between adjacent submodules had the best performance. The performance of the DNFS can be improved by distributing the features with high correlation with the output to each submodule. Given the results of the current study, it is expected that the DNFS will be used to solve general high-dimensional regression problems efficiently with high accuracy and better interpretability.
引用
收藏
页数:19
相关论文
共 40 条
[1]   Wavelet-adaptive neural subtractive clustering fuzzy inference system to enhance low-cost and high-speed INS/GPS navigation system [J].
Abdolkarimi, E. S. ;
Mosavi, M. R. .
GPS SOLUTIONS, 2020, 24 (02)
[2]   The prediction of longitudinal dispersion coefficient in natural streams using LS-SVM and ANFIS optimized by Harris hawk optimization algorithm [J].
Azar, Naser Arya ;
Milan, Sami Ghordoyee ;
Kayhomayoon, Zahra .
JOURNAL OF CONTAMINANT HYDROLOGY, 2021, 240
[3]   Improved adaptive neuro-fuzzy inference system based on modified glowworm swarm and differential evolution optimization algorithm for medical diagnosis [J].
Balasubramanian, Kishore ;
Ananthamoorthy, N. P. .
NEURAL COMPUTING & APPLICATIONS, 2021, 33 (13) :7649-7660
[4]   Forecasting energy consumption using ensemble ARIMA-ANFIS hybrid algorithm [J].
Barak, Sasan ;
Sadegh, S. Saeedeh .
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS, 2016, 82 :92-104
[5]   Exploration of Principal Component Analysis: Deriving Principal Component Analysis Visually Using Spectra [J].
Beattie, J. Renwick ;
Esmonde-White, Francis W. L. .
APPLIED SPECTROSCOPY, 2021, 75 (04) :361-375
[6]   Prediction of Heart Attack Risk Using GA-ANFIS Expert System Prototype [J].
Begic Fazlic, Lejla ;
Avdagic, Aja ;
Besic, Ingmar .
PROCEEDINGS OF THE 12TH INTERNATIONAL CONFERENCE ON WEARABLE MICRO AND NANO TECHNOLOGIES FOR PERSONALIZED HEALTH (PHEALTH 2015), 2015, 211 :292-294
[7]   Adaptive particle swarm optimized fuzzy algorithm to predict water table elevation [J].
Bisht, Dinesh C. S. ;
Jain, Shilpa ;
Srivastava, Pankaj Kumar .
INTERNATIONAL JOURNAL OF MODELING SIMULATION AND SCIENTIFIC COMPUTING, 2019, 10 (06)
[8]   New insights into soil temperature time series modeling: linear or nonlinear? [J].
Bonakdari, Hossein ;
Moeeni, Hamid ;
Ebtehaj, Isa ;
Zeynoddin, Mohammad ;
Mahoammadian, Abdolmajid ;
Gharabaghi, Bahram .
THEORETICAL AND APPLIED CLIMATOLOGY, 2019, 135 (3-4) :1157-1177
[9]   Prediction and Analysis of Literature Loan Circulation in University Libraries Based on RBF Neural Network Optimized Model [J].
Chen, Xia ;
Yang, Wanqin .
AUTOMATIC CONTROL AND COMPUTER SCIENCES, 2020, 54 (02) :139-146
[10]   Optimize TSK Fuzzy Systems for Classification Problems: Minibatch Gradient Descent With Uniform Regularization and Batch Normalization [J].
Cui, Yuqi ;
Wu, Dongrui ;
Huang, Jian .
IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2020, 28 (12) :3065-3075