Data stored in repositories are rapidly growing in terms of instances represented with multiple attributes/dimensions. To represent characteristics of an instance, mixed type attributes are used. Banking System is one of the areas which store information of bank customers in multiple dimensions. Principal Component Analysis (PCA) is a Dimensionality reduction technique in Data Mining used to transform the attributes of a dataset to a lesser dimensional space. Classification is a Supervised Machine Learning technique used to distinguish the instances of a dataset into a number of classes. In this work, we have analyzed the Bank Marketing dataset containing 1000 instances of bank clients represented with 17 attributes, with a class label as the last attribute. Principal Components (PCs) are generated from the input dataset by applying PCA on mixed attributes. A Deep Neural Network classifier is built by applying Backpropagation on the PCs. Experimental results show that our proposed PCA mixed Deep Neural Network classifier outperforms existing classifiers in terms of accuracy.

Banking System, Mixed Data, PCA, Classification, Backpropagation, Deep Neural Network

[1] Dunhamm M. H., “Data Mining: Introductory and Advanced Topics”, Pearson Education, India, 2006.
[2] Han J. and Kamber M., “Data Mining Concepts and Techniques”, Morgan Kauffmann Publishers, India, 2006.
[3] Lin C. and Yan F., “The Study on Classification and Prediction for Data Mining”, Seventh Int’l Conf. on Measuring Technology and Mechatronics Automation, 2015.
[4] Zhang P.G., “Neural networks for classification: a survey”, IEEE Trans. Systems, Man, and Cybernetics, Part C (Applications and Reviews), Vol. 30, Issue 4, Nov 2000.
[5] Gao T., Li X., Chai Y. and Tang Y., “Deep Learning with Stock Indicators and two-dimensional Principal Component Analysis for closing price prediction system”, Seventh IEEE Int’l Conf. Software Engineering and Service Science, IEEE, Aug 2016.
[6] Shi H. and Liu X., “Application on Stock Price Prediction of Elman neural networks based on Principal Component Analysis Method”, 11th Int’l Conf. Wavelet Actiev Media Technology and Information Processing, IEEE, Dec 2014.
[7] Ming C.T.J., Noor M.N, Rijal M.O., Kassim M.R and Yunus A., “Lung Disease Classification Using Different Deep Learning Architectures and Principal Component Analysis”, 2nd Int’l Conf. BioSignal Analysis, Processing and Systems, IEEE, July 2018.
[8] Feng W., Zhao Y. and Deng J., “Application of SVM Based on Principal Component Analysis to Credit Risk Assessment in Commercial Banks”, WRI Global Congress on Intelligent Systems, IEEE, May 2009.
[9] Min Z., “Credit Risk Assessment Based on Fuzzy SVM and Principal Component Analysis”, Int’l Conf. Web Information Systems and Mining, IEEE, Nov 2009.
[10] Ioniţă I. and Şchiopu D., “Using Principal Component in Loan Granting”, Seria Matematică – Informatică – Fizică, Vol. LXII, pp. 88-96, 2010.
[11] https://pbpython.com/categorical-encoding.html
[12] https://en.wikipedia.org/wiki/Principal_component_analysis
[13] https://en.wikipedia.org/wiki/Backpropagation
[14] https://archive.ics.uci.edu/ml/datasets/bank+marketing
[15] https://www.jetbrains.com/pycharm/
[16] Naraei P., Abhari A. and Sadeghian A., “Application of Multilayer Perceptron Neural Networks and Support Vector Machines in Classification of Healthcare Data”, Future Technologies Conference, IEEE, Dec. 2016.
[17] Abd-Alsabour N., “A Review on Evolutionary Feature Selection”, European Modelling Symposium, IEEE, Oct 2014.
[18] Holden N. and Freitas A.A., “A hybrid particle swarm/ant colony algorithm for the classification of hierarchical biological data”, Proc. Swarm Intelligence Symposium, IEEE, June 2005.
[19] Fernandes M., “Data Mining: A Comparative Study of its Various Techniques and its Process”, International Journal of Scientific Research in Computer Science and Engineering, Vol. 5, Issue 1, Feb. 2017, pp.19-23.
[20] Ghuse N., Pawar P. and Potgantwar A., “An Improved Approach For Fraud Detection In Health Insurance Using Data Mining Techniques”, IJSRNSC, Vol. 5, Issue 5, June 2017.