2 results
Search Results
Now showing 1 - 2 of 2
Item Open Access INTRODUCE PREDICTIVE ANALYTICS USING THE NEXT BEST ACTION (NBA) MODELS INTO THE BANKING SYSTEM(СДУ хабаршысы - 2020, 2020) Dauylov S. ; Bogdanchikov A.Abstract. NBA - is an approach in which each client is initially considered purely individual. It has a close correlation with Predictive analysis. Predictive or prognostic analytics is a set of techniques and methods for analyzing data to build a forecast of future events. The banking system is currently using the method to obtained certain business results from its customers and has increased loyalty, increased income, found new growth points, etc. The classical model of marketing was rather different, it repelled from its existing product line and its parameters. But new models repels from customer’s inclination to purchase a particular product. The aim of this project is to investigate the field of deposit accounts of the banking system by using NBA approach and to show the benefits and possible outcomes. This approach was tested on various aspects of the banking system and showed a number of solutions which can predict the probability of a customer to create a term deposit account.Item Open Access FORECASTING OIL PRODUCTION USING LSTM NETWORKS CONFINED TO DECLINE(СДУ хабаршысы - 2020, 2020) Zhumekeshov A. ; Bogdanchikov A.Abstract. Natural resources are limited and very important in our industrial life and development. Oil is considered as the black gold and it is included in hundreds of industrial fields. Therefore, forecasting future oil production performance is an important aspect for oil industry. In this study, we proposed improvements to the existing deep learning model in order to overcome limitations associated with the original model. For evaluation purpose, proposed and original deep learning models were applied on a real case oil production data. The empirical results show that the proposed adjustments to the existing deep learning model achieves better forecasting accuracy.