13-11-2014, 10:04 AM
Abstracts: Due to rapid advancement in the electronic commerce technology, the use of ATM cards has dramatically increased. As ATM card becomes popular mode of payment for both online as well regular purchases, cases of fraud associated with it also rising. In this model the sequence of operations in ATM card transaction processing using Hidden Markov Model as well as Bluetooth based mobile security model and show how it can be used for the detection of frauds. To determine parameters of HMM, it is first necessary to make rough guess at what they might be. A HMM is a statistical model using a Markov process with observed state, thus is clearly a dynamic Bayesian network of statistical assumption technique. Mobile device induced distributed security environment is proposed for meeting pass code challenge to complicate security by making the ATM payment frauds impossible. If an incoming ATM card transaction is not accepted by the trained HMM with sufficiently high probability, it is considered to be fraudulent, Assuring that genuine transactions are not rejected. Hidden state is not directly visible, but output dependent on the state is visible. Some relevant facts that some applications like speech recognition system, DNA analysis are already being used effectively which uses the concepts related to pattern generation and recognition. These applications use strongly build Hidden Markov Model that intern is totally based on Compound decision making theory from probability and statistics. Thus all such applications use the algorithms and concepts that finally help in making the decision based on multiple factors. HMM uses many of such like Baum-Welch, FW-BW, Viterbi, K-Means etc.