Mitigating Financial Fraud Using Data Science - “A Case Study on Credit Card Frauds”

In the early years of fast and active development of 80s and 90s, financial frauds happened to be pretty simple. They were not more than duplicity of forged cheques, draining peoples or investors' money through fake company formations, and minutest scrutinization of the loan documents etc. These were the only means for the fraudsters to play around. It is only after the advancement and adoption of desktop culture we have witnessed the whole new age of cybercrime and digital frauds. Investors, retailers, businesses and corporates none were spared and were hard hit. Since then, financial frauds have become intimidating for businesses and especially for the banks across the globe. With the advent and continuous advancement of technology it has further complicated the ways and means for the fraudsters ending up into catastrophic consequences. As data is growing many related connected challenge are also increasing. With the help of data science various aspect of data can be analyzed, various pattern of accessing the data can be understood, which can eventually help to understand the probability of risk associated with various pattern of storing / accessing / retrieving the data. This paper also presents the an analysis on open source dataset, taken from Kaggel, for the data analysis by using logistic regression, and the results of which are measure with confusion matrix, which provides more clear understand of the dataset.