• Aditya Mule
• Ayush Saraf
• Sargam Pandey
• Sakshi Bhosale

Credit card transactions occurs in a tremendous rate so keep track on these transactions manually to categorize fraudulent transactions and non-fraudulent transactions is practically very hard, so there must be a mechanism that will do it automatically. Many information mining techniques had been proposed for fraud detection. In this paper, a new idea of core structured collection gaining knowledge of structures has been proposed by staking a group of version, a deep sequential learning version and another top layered group classifier in right order. Models in this structure have been cleared to be very efficient in scenarios like fraud detection, where the information sequence is made up of vectors with complex interconnected features. Fraud detection has become the vital activity to reduce fraud impact on service quality, costs and reputation of a company or institute. Traditional anti-fraud method relying on manual audit is unable to deal with explosively growing information data. Meanwhile, criminals are keeping on finding new tricks by avoiding known rules to commit fraud actions. Financial institutions are struggling to find more intelligent methods for detecting fraud events, with the goal of reducing fraud losses as much as possible.

Recurrent Neural Network, Classification, LSTM.

"REAL TIME FRAUD PREDICTION USING ADVANCED NEURAL NETWORK", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 4, page no.389-391, April-2019, Available :http://www.jetir.org/papers/JETIR1904671.pdf

"REAL TIME FRAUD PREDICTION USING ADVANCED NEURAL NETWORK", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 4, page no. pp389-391, April-2019, Available at : http://www.jetir.org/papers/JETIR1904671.pdf