Asymmetric Support Vector Regression Approach in ATM Cash Replenishment Optimization

ATM cash replenishment problem is quite important in banking and finance sector. As well as choosing the forecast model giving the smallest forecast error ratio for the solution of this problem, finding the optimization model giving the minimum replenishment costs has importance. In this study, a new asymmetrical forecast model and an optimization model running integrated with the forecast model, in other words, an optimization model which reduces the two stage forecast and optimization process to a single step is proposed. At the same time, a comparison of costs with the other forecast models is performed.