Research And Implementation Of Fine-grained Operation Modeling Method For Bank Self-service Terminals

Postal Savings Bank of China,Shenzhen Branch improves the efficiency of self-service equipment through the establishment of self-service equipment management platform,and reduces cost of equipment operation and capital occupancy by scientific rational prediction of self-service equipment in the case of characteristic value.The paper begins with the establishment of self-service equipment management platform system,illustrates the attribute checking and application relevant to self-service equipment,and ultimately outputs the operational data of the equipment in a digital and quantized way,thus providing intelligent support for making production and operating decision-making.Its main research content and overview is as follows:Over recent years,with the extensive distribution of self-service equipment,clearing apparatus and cash replenishment is a significant indicator link in operational management.At current,it's challenged by a prominent problem.On the one hand,due to the limitation on transport capacity and manpower,cash replenishment has weakness,which always lack of cash;on the other hand,there should be no redundancy of stock of notes for capital cost limitation.In response to the problems stated above,the paper focuses on this indicator,sets up a self-service equipment operation platform on a basis of research results,in a bid to realize autonomous management and account statement collection for providing prediction with data sources,thus finally achieving the prediction of the current amount of cash replenishment,which is the main research work and results is as follows.First,in this paper,we put forward a regression method based on urban thermodynamics chart for accuracy in the amount of cash replenishment,and verify the algorithm of production data in practice,while solved the standard algorithm excludes the consideration of data isolation and service forecast value susceptible to multiple factors.Second,in this paper,we improve finite difference autoregressive model to forecast the amount of cash replenishment based on mobile payment factors.To improve forecast accuracy,we introduce the notion of coefficient,and restructure the formula according to the change of mobile payment transact ion in combination with the predictive coefficient generated through actual production data.With improved content correlation and forecast accuracy,we can further improve the performance of the model.In the experiment,a forecast effect better than trad itional finite difference autoregressive model is obtained.Finally,this paper put forward the integrative regression learning method by multiple influence factors.Through the learning of a group of differentiated weights for prediction of diversified classifier portfolio,the regression comparison experiment and ensemble learning experiment suggest that the proposed prediction model has favorable effect.Based on it,a bank self-service equipment production and operational system is designed and realized,the prediction is made based on account statement data collection,and the equipment is put into operation,thus realizing the delicacy management for bank self-service terminal.