An Evaluation of Forecasting Precision with Forecasting Method Concerning Water Demand for A Prefecture  
Author Yukio Maruyama

 
Co-Author(s) Yutaka Shirai;  Kazuo Kihara

 
Abstract Forecasting the water demands is very important for pure-water reservoirs. Because the amount of water supply to the water supply stations cannot be changed frequently according to the variation of water demands. In our studies, we proposed three kinds of forecasting method and one kind of correcting method for water demand using data of five prefectures or town in Japan. The points of these forecasting methods are following three matters. 1) Based on a multiple regression analysis. 2) Using multiple year's data of the value of water supply per day. 3) Correcting the obtained multivariate regression formula by single regression analysis. In previous study, it was confirmed that our proposed forecasting method can be applied to forecasting the duration of COVID-19. The purpose of this paper is to examine parameter combinations that improve the forecasting precision in our proposed forecasting method. As the result of this examination, different parameter combination is obtained than previous studies. We hope that the result of this examination can be applied as one of the parameter selection methods in other forecasting methods using deep learning.

 
Keywords Water demand, statistical model, multiple regression analysis, forecasting precision, estimation of parameters
   
    Article #:  RQD2025-96
 

Proceedings of 30th ISSAT International Conference on Reliability & Quality in Design
August 6-8, 2025