A classification algorithm with Linear Discriminant Analysis and Axiomatic Fuzzy Sets

Wenjuan Jia; Yingjie Deng; Chenyang Xin; Xiaodong Liu; Witold Pedrycz

doi:10.3934/mfc.2019006

Article Contents

2019, Volume 2, Issue 1: 73-81. Doi: 10.3934/mfc.2019006

This issue Previous Article Online optimization for residential PV-ESS energy system scheduling Next Article

A classification algorithm with Linear Discriminant Analysis and Axiomatic Fuzzy Sets

1.
School of Control Science and Engineering, Faculty of Electronic Information and Electrical Engineering, Dalian University of Technology, Dalian, China
2.
Department of Electrical and Computer Engineering, University of Alberta, Edmonton, T6G 2G7, Canada

^* Corresponding author: Xiaodong Liu
^* Corresponding author: Xiaodong Liu

Published: March 2019

The work is supported by National Natural Science Foundation of China under grants 61673082 and 61533005.

Abstract / Introduction Full Text(HTML) Figure(7) / Table(2) Related Papers Cited by

Abstract

In exploratory data mining, most classifiers pay more attention on the accuracy and speed of learned models, but they are lacking of the interpretability. In this paper, an interpretable and comprehensible classifier is proposed based on Linear Discriminant Analysis (LDA) and Axiomatic Fuzzy Sets (AFS). The algorithm utilizes LDA to extract features with the largest inter-class variance. Besides, the proposed approach aims to explore a transformation from the selected feature space to a semantic space where the samples in the same class are made as close as possible to one another, whereas the samples in the different class are as far as possible from one another. Moreover, the descriptions of each class can be obtained by the proposed approach. When compared with well-known classifiers such as LogisticR, C4.5Tree, SVM and KNN, the proposed method not only can achieve better performance in terms of accuracy but also has the capability of interpretability and comprehension.

Keywords:

Mathematics Subject Classification: Primary: 03B52, 03E72, 28E10, 94D05.

Citation:

Full Text(HTML)

Figure 1. The proposed classifier flow chart

Download: Full-size image PowerPoint slide

Figure 2. Samples in abstract features space

Download: Full-size image PowerPoint slide

Figure 3. The membership degree of $ m_1 $ on all data

Download: Full-size image PowerPoint slide

Figure 4. The membership degree of description Class 1

Download: Full-size image PowerPoint slide

Figure 5. The membership degree of description Class 2

Download: Full-size image PowerPoint slide

Figure 6. The membership degree of description Class 3

Download: Full-size image PowerPoint slide

Figure 7. The membership degree of three class descriptions on all samples

Download: Full-size image PowerPoint slide

Table 1. The maximum, average and minimum of each abstract feature

feature	$ f_1 $	$ f_2 $
minimum	-1.73	-2.36
average	0.00	0.00
maximum	1.61	2.44

| Show Table

DownLoad: CSV

Table 2. The experiments results-accuracy rates(standard deviation)

dataset	LogisticR	C4.5Tree	SVM	KNN	Our method
wine	0.9556$ \pm $0.0032	0.9364$ \pm $0.0127	0.7900$ \pm $0.0118	0.7074$ \pm $0.0141	0.9666$ \pm $0.0005
iris	0.9593$ \pm $0.0021	0.9520$ \pm $0.0053	0.9826$ \pm $0.0071	0.9647$ \pm $0.0032	0.9867$ \pm $0.0000
heart	0.8399$ \pm $0.0038	0.7544$ \pm $0.0230	0.6918$ \pm $0.0181	0.6604$ \pm $0.0118	0.8407$ \pm $0.0000
breast_C	0.7220$ \pm $0.0209	0.7142$ \pm $0.0209	0.6034$ \pm $0.0241	0.5405$ \pm $0.0118	0.7753$ \pm $0.0038
seeds	0.9228$ \pm $0.0020	0.9286$ \pm $0.0089	0.9271$ \pm $0.0055	0.8828$ \pm $0.0088	0.8590$ \pm $0.0093
USD	0.7309$ \pm $0.0052	0.9321$ \pm $0.0090	0.9510$ \pm $0.0017	0.8247$ \pm $0.0133	0.7912$ \pm $0.0090
column_2c	0.8258$ \pm $0.0030	0.8067$ \pm $0.0205	0.8625$ \pm $0.0031	0.8280$ \pm $0.0064	0.7697$ \pm $0.0040
caesarian	0.6741$ \pm $0.0189	0.5263$ \pm $0.0341	0.6551$ \pm $0.0161	0.5589$ \pm $0.0326	0.7253$ \pm $0.0125
immunotherapy	0.7973$ \pm $0.0139	0.8073$ \pm $0.0262	0.7897$ \pm $0.0000	0.7235$ \pm $0.0208	0.7617$ \pm $0.0091
SHS2015	0.5572$ \pm $0.0133	0.6126$ \pm $0.0133	0.6443$ \pm $0.0213	0.5445$ \pm $0.0561	0.6498$ \pm $0.0028

| Show Table

DownLoad: CSV

Related Papers

Cited by

References

[1]	W. Bi, M. Cai, M. Liu and et al., A Big Data Clustering Algorithm for Mitigating the Risk of Customer Churn, IEEE Transactions on Industrial Informatics, 12 (2016), 1270-1281.
[2]	A Das and A. Desarkar, Decision tree-based analytics for reducing air pollution, Journal of Information & Knowledge Management, 17 (2018), 1850015.
[3]	K. J. Dsouza and Z. Ansari, Experimental exploration of support vector machine for cancer cell classification, IEEE International Conference on Cloud Computing in Emerging Markets, 2017, 29–34.
[4]	D. Dua and E. Karra Taniskidou, UCI Machine Learning Repository, http://archive.ics.uci.edu/ml. Irvine, CA: University of California, School of Information and Computer Science, 2017.
[5]	L. I. Jinmeng, Y. Lin and T. Zhu, k-Nearest Neighbor Classification Algorithm Based on Hubness and Class Weighting, Computer Engineering, 44 (2018), 248-252.
[6]	F. Khozeimeh, R. Alizadehsani, M. Roshanzamir, A. khosravi and et al., An expert system for selecting wart treatment method, Computer in Biology and Medicine, 81 (2017), 167-175.
[7]	F. Khozeimeh, F. Jabberi Azad and et al., Intralesional immunotherapy compared to cryotherapy in the treatment of warts, International Journal of Dermatology, 56 (2017), 474-478.
[8]	Q. Li, Y. Ren, L. Li and W. Liu, From low-level geometric features to high-level semantics: An axiomatic fuzzy set clustering approach, Journal of Intelligent & Fuzzy Systems, 31 (2016), 775-786. doi: 10.3233/JIFS-169009.
[9]	Q. Li, Y. Ren, L. Li and et al., Fuzzy based affinity learning for spectral clustering., Pattern Recognition, 60 (2016), 531-542.
[10]	Z. Li, Q. Zhang and X. Duan, et al., A novel semantic approach for multi-ethnic face recognition, International Journal of Pattern Recognition & Artificial Intelligence, 32 (2017), 15.
[11]	X. Liu, W. Jia and Y. Wang, et al., Knowledge discovery and semantic learning in the framework of axiomatic fuzzy set theory, Wires Data Mining And Knowledge Discovery, 8 (2018), e1268. doi: 10.1002/widm.1268.
[12]	X. Liu, K. Zhu and H. Z. Huang, The representations of fuzzy concepts based on the fuzzy matrix theory and the AFS theory, IEEE International Symposium on Intelligent Control, 2003.
[13]	X. Liu, The fuzzy theory based on afs algebras and afs structure, Journal of Mathematical Analysis and Applications, 217 (1998), 459-478. doi: 10.1006/jmaa.1997.5718.
[14]	X. Liu, W. Wang, T. Chai and et al., Approaches to the representations and logic operations of fuzzy concepts in the framework of axiomatic fuzzy set theory I, Information Sciences, 177 (2007), 1007-1026. doi: 10.1016/j.ins.2006.07.011.
[15]	X. Liu, The fuzzy sets and systems based on AFS structure, EI algebra and EII algebra, Fuzzy Sets & Systems, 95 (1998), 179-188. doi: 10.1016/S0165-0114(96)00298-9.
[16]	X. Liu, X. Feng and W. Pedrycz, Extraction of fuzzy rules from fuzzy decision trees: An axiomatic fuzzy sets (AFS) approach, Data & Knowledge Engineering, 84 (2013), 1-25. doi: 10.1016/j.datak.2012.12.001.
[17]	Y. Ren, M. Song and X. Liu, New approaches to the fuzzy clustering via AFS theory, International Journal of Information and Systems Sciences, 3 (2007), 307-325.
[18]	Y. Ren, Q. Li, W. Liu and et al., Semantic facial descriptor extraction via Axiomatic Fuzzy Set, Neurocomputing, 171 (2016), 1462-1474.
[19]	A. Swami and R. Jain, Scikit-learn: Machine learning in python, Journal of Machine Learning Research, 12 (2012), 2825-2830.
[20]	W. Wang and X. Liu, Fuzzy forecasting based on automatic clustering and axiomatic fuzzy set classification, Information Sciences, 294 (2015), 78-94. doi: 10.1016/j.ins.2014.09.027.
[21]	I. H. Witten, E. Frank, M. A. Hall and C. J. Pal, Data Mining: Practical Machine Learning Tools and Techniques (Fourth Edition), Elsevier, Singapore, 2017.
[22]	X. Zhang, J. M. Chiou and Y. Ma, Functional prediction through averaging estimated functional linear regression models, Biometrika, 105 (2018), 945-962. doi: 10.1093/biomet/asy041.