Paper Title
ATM Replenishment Forecasting with Support Vector Machine – A Case Study

Abstract
In the cash management process of banks, Automated Teller Machine (ATM) hold an important role. The correct ATM replenishment policies is directly dependent to the correct forecasting of the cash transactions such as withdrawal and deposits. The money in the ATMs are deprived from interest, however the lack of necessary amount of money might yield to the client loss. Therefore, the forecasting studies concerning the ATMs is an important topic that investigated in the literature. In this study, a machine learning model by using Support Vector Regressor (SVR) algorithm is applied to predict daily cash withdrawals of 5 ATMs considering different factors. Keywords - Cash Management, Time Series Prediction, Machine Learning, SVM