ATM is one of the most pressing issues in today's banking system. The popularity of a bank will decline if an ATM has a lack of cash, and this will lead to increased costs for the bank and a decrease in customer use of ATMs. In order to ensure which neither a consumer's transaction is refused due to the ATM being out of cash, nor the bank's profit potential is squandered, each ATMs cash must be well stocked. Managing the quantity of currency in an ATM is critical to any bank's ability to serve its customers. For the most part, banks use third-party cash management firms to keep ATMs topped up on a regular basis. They're doing a study to see whether analysis of the data and Machine Learning (ML) can be used to supplement the present system's mathematical capabilities. Hence, this paper provides a survey on researches of predicting the proper amount of ATM cash replenishment to ensure that the bare minimum of cash is always present until the next refill. ATMs daily cash limit is relatively a time series phenomenon but it is has difficulty in prediction. There'll be no client unhappiness as a result of an ATM cash out issue is addressed by employing a data driven technique to estimate the proper quantity for each ATM or set of ATMs, an ATM replenishment prediction machine learning approach.
Key words: ATM cash prediction, Machine learning approaches, Time series, Cash demand prediction, Replenishment amount, Cash out.
|
