Credit Card Fraud Analysis using Robust Space Invariant Artificial Neural Networks (RSIANN)
S. Deepika1, S. Senthil2
1S. Deepika, Research scholar (Rg. No: R16PCS09), REVA University, Asst. Professor, Anurag group of Institutions (Autonomous), Hyderabad.
2S. Senthil, Professor & Director- School of Computer Science and Applications, Reva University, Bangalure.
Manuscript received on 04 March 2019 | Revised Manuscript received on 11 March 2019 | Manuscript published on 30 July 2019 | PP: 6413-6417 | Volume-8 Issue-2, July 2019 | Retrieval Number: B2315078219/19©BEIESP | DOI: 10.35940/ijrte.B2315.078219
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: One of the impact factor for any organizations or banks revenue and service quality is credit card fraud activities. Hence, need of efficient approach for detect early potential fraud and/or prevent them. In this paper, we considered pre-processing and used deep convolution neural network called as Space Invariant Artificial Neural Networks for classifying fraudsters. Available Credit card fraud dataset may not have sufficient information hence need pre-processing. The proposed approach has pre-processing phrase to make as robust. This approach used leverage layers and suitable tuning parameters for getting good classification accuracy. In neural network applications, choosing of tuning parameters and model selection has great role in solving the problems. We have done careful analysis and selected leverage layers and corresponding parameter values. The proposed architecture tested with all possible tuning parameters to evaluate the performance on pre-processed credit card fraud records. We found the proposed robust SIANN (RSIANN) is outperformed other state-of-art machine learning (ML) algorithms (Support vector machine (SVM), random forest (RF), Navie bayes and deep convolution neural network (DCNN) in terms of accuracy (85%). Thus, this model analyses the transaction and decide it fraud or not.
Index Terms: Credit Card Fraud, CNN, Pre-Processing, Machine Learning.
Scope of the Article: Machine Learning