International Journal of Computer Science & Network Security

International Journal of Computer Science & Network Security (국제컴퓨터통신보호논문지학회)
권호 표지
DOI QR코드

DOI QR Code

A Multi-Class Classifier of Modified Convolution Neural Network by Dynamic Hyperplane of Support Vector Machine

  • Nur Suhailayani Suhaimi (Faculty of Computer and Mathematical Sciences, University of MARA Technology) ;
  • Zalinda Othman (Center for Artificial Intelligence Technology, Faculty of Information Science & Technology, The National University of Malaysia) ;
  • Mohd Ridzwan Yaakub (Center for Artificial Intelligence Technology, Faculty of Information Science & Technology, The National University of Malaysia)
  • Received : 2023.11.05
  • Published : 2023.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we focused on the problem of evaluating multi-class classification accuracy and simulation of multiple classifier performance metrics. Multi-class classifiers for sentiment analysis involved many challenges, whereas previous research narrowed to the binary classification model since it provides higher accuracy when dealing with text data. Thus, we take inspiration from the non-linear Support Vector Machine to modify the algorithm by embedding dynamic hyperplanes representing multiple class labels. Then we analyzed the performance of multi-class classifiers using macro-accuracy, micro-accuracy and several other metrics to justify the significance of our algorithm enhancement. Furthermore, we hybridized Enhanced Convolution Neural Network (ECNN) with Dynamic Support Vector Machine (DSVM) to demonstrate the effectiveness and efficiency of the classifier towards multi-class text data. We performed experiments on three hybrid classifiers, which are ECNN with Binary SVM (ECNN-BSVM), and ECNN with linear Multi-Class SVM (ECNN-MCSVM) and our proposed algorithm (ECNNDSVM). Comparative experiments of hybrid algorithms yielded 85.12 % for single metric accuracy; 86.95 % for multiple metrics on average. As for our modified algorithm of the ECNN-DSVM classifier, we reached 98.29 % micro-accuracy results with an f-score value of 98 % at most. For the future direction of this research, we are aiming for hyperplane optimization analysis.

Keywords

Acknowledgement

We are very grateful to Universiti Teknologi MARA for supporting this research. We sincerely thank our research lab, Data Mining and Optimization (DMO) of Fakulti Teknologi dan Sains Maklumat (FTSM), Universiti Kebangsaan Malaysia, for the expert knowledge sharing. Thanks to Universiti Kebangsaan Malaysia and the Ministry of Higher Education for providing Fundamental Research Grant Scheme (FRGS) code FRGS/1/2021/ICT06/UKM/02/1 for this research funding.

References

  1. Ahmad, I.S., Bakar, A.A. and Yaakub, M.R.: A Survey on Machine Learning Techniques in Movie Revenue Prediction. Springer Nature Computer Science, 1(235), (2020). 
  2. Al-Aidaroos, K.M., Bakar, A.A. and Othman, Z.: Naive Bayes Variants in Classification Learning. IEEE Xplore, International Conference on Information Retrieval & Knowledge Management, 276-281 (2010). 
  3. Ansari, Z.A., Sattar, S.A. and Babu, A.V. A fuzzy neural network based framework to discover user access patterns from web log data. Journal of Advances in Data Analysis and Classification. Springer-Verlag Berlin Heidelberg. (2015). 
  4. Awwalu, J., Bakar, A.A. and Yaakub, M.R.: Hybrid Ngram Model using Naive Bayes for Classification of Political Sentiments on Twitter. Springer Nature Neural Computing and Applications, 31, 9207-9220 (2019). 
  5. Barman, D. and Chowdhury, N. A novel semi supervised approach for text classification. International Journal of Information Technology. DOI: https://doi.org/10.1007/s41870-018-0137-9. (2018). 
  6. Benitez-Pena, S., Blanquero, R., Carrizosa, E. and Ramirez-Cobo, P. On support vector machines under a multiple-cost scenario. Journal of Advances in Data Analysis and Classification. Springer Nature 2018: Springer-Verlag GmbH Germany. (2018). 
  7. Berrendero, J. R. and Carcamo, J. Linear Components of Quadratic Classifiers. Journal of Advances in Data Analysis and Classification. Springer Nature 2018: Springer-Verlag GmbH Germany. (2018). 
  8. Blanco, V., Japon, A. and Puerto, J. Optimal arrangements of hyperplanes for SVM-based multiclass classification. Journal of Advances in Data Analysis and Classification. Springer Nature 2019: Springer-Verlag GmbH Germany. (2019). 
  9. Burg, G.J.J.V-D. and Groenen. P.J.F.: GenSVM: A Generalized Multiclass Support Vector Machine. Journal of Machine Learning Research, 17, 1-42, (2016).
  10. Cappozzo, A., Greselin, F. and Murphy, T.B. A robust approach to model-based classification based on trimming and constraints; Semi-supervised learning in presence of outliers and label noise. Journal of Advances in Data Analysis and Classification. Springer Nature 2019: Springer-Verlag GmbH Germany. (2019). 
  11. Chaitra, P.C. and Kumar R.S.: Review of Multi-Class Classification Algorithms. International Journal of Pure and Applied Mathematics 118(14), 17-26 (2018). 
  12. Ghareb, A.S., Bakar, A.A. and Hamdan, A.R.: Hybrid Feature Selection based on enhanced Genetic Algorithm for Text. Elsevier Journal for Expert Systems with Applications 18, 21-44 (2015). 
  13. Grandini, M., Bagli, E. and Visani, G.: Metrics for Multi-Class Classification: An Overview. Computer Science, Mathematics, arXiv: 221112671, (2020). 
  14. Horn, D., Demircioglu, A., Bischl, B., Glasmachers, T. and Weihs, C. A comparative study on large scale kernelized support vector machines. Journal of Advances in Data Analysis and Classification. Springer Springer-Verlag Berlin Heidelberg (2016). 
  15. Jayadeva, Khemchandani, R. & Chandra, S. Twin Support Vector Machines for Pattern Classification, IEEE Trannsaction on Pattern Analysis and Machine Intelligence, 29 (5): 905-10. (2007)  https://doi.org/10.1109/TPAMI.2007.1068
  16. Ju, X., Tian Y., Liu, D. and Qi, Z. Nonparallel Hyperplanes Support Vector Machine for Multi-class Classification. Elsevier: International Conference On Computational Science (51), pp. 1574-1582. (2015) 
  17. Kumar, A., Dabas, V. and Hooda, P. Text classification algorithms for mining unstructured data: a SWOT analysis. International Journal of Information Technology. DOI: https://doi.org/10.1007/s41870-017- 0072-1. (2017). 
  18. Kuo, K.M., Yalley, P., Kao, Y. and Huang, C.H.: A Multi-Class Classification Model for Supporting the Diagnosis of Type II Diabetes Melitus. PeerJ, 8, e9920, (2020). 
  19. Lausser, L., Schmid, F., Schirra, L-R., Wilhelm, A.F.X. and Kestler, H.A. Rank-based Classifier for Extremely High-dimensional Gene Expression Data. Journal of Advances in Data Analysis and Classification. SpringerVerlag Berlin Heidelberg. (2016). 
  20. Li, L., Zhao, K., Sun, R., Gan, J., Yuan, G. and Tong, L. Parameter-Free Extreme Learning Machine for Imbalanced Classification. Journal of Neural Processing Letters. Springer Science+Business Media, LLC, part of Springer Nature (2020 ). 
  21. Matthew, B. and Sohini, R.: A Generalized Flow for Multi-class and Binary Classification Tasks: An Azure ML Approach. IEEE International Conference on Big Data (Big Data), 1728-1737 (2015). 
  22. Pei, H., Lin, Q., Yang, L. and Zhong, P. A novel semisupervised support vector machine with asymmetric squared loss . Journal of Advances in Data Analysis and Classification. Springer Nature 2020: Springer-Verlag GmbH Germany. (2020). 
  23. Rathgamage, D. and Duleep, P.W.: Multiclass Classification Using Support Vector Machines. Electronic Theses and Dissertations, 1845, (2018). 
  24. Rathgamage, D. and Iacob, I.E.: DCSVM: Fast Multiclass Classification using support Vector Machines. Springer: International Journal of Machine Learning and Cybernetics, (2019). 
  25. Saigal, P. and Khanna, V.: Multi-category News Classification using Support Vector Machine based Classifiers. Springer Nature Applied Sciences, 2 (3), (2020). 
  26. Salter-Townshend, M. and Murphy T.B. Mixtures of biased sentiment analysers. Journal of Advances in Data Analysis and Classification. Springer-Verlag Berlin Heidelberg. (2013). 
  27. Sueno, H.T., Gerardo, B.D. and Medina R.P.: Multiclass Document Classification using Support Vector Machine (SVM) Based on Improved Naive Bayes Vectorization Technique. International Journal of Advanced Trends in Computer Science and Engineering, 9 (3), 3937-3944, (2020). 
  28. Yaakub, M.R., Latiffi, M.I.A. and Zaabar, L.S.: A Review on Sentiment Analysis Techniques and Applications. IOP Conference Series: Materials Science and Engineering, 55, (2019). 
  29. Yang, L. and Wang, L. A class of semi-supervised support vector machines by DC programming. Journal of Advances in Data Analysis and Classification. Springer-Verlag Berlin Heidelberg. (2013). 
  30. Zhao, X., Barber, S., Taylor, C. C., and Milan, Z. Interval forecasts based on regression trees for streaming data. Journal of Advances in Data Analysis and Classification. Springer Nature 2019: SpringerVerlag GmbH Germany. (2019). 
  31. Zhu, W. and Zhong, P. Minimum Class Variance SVM+ for data classification. Journal of Advances in Data Analysis and Classification. Springer Springer-Verlag Berlin Heidelberg (2015).