Improvement and Evaluation of Estimation of Time Series Data of Daily Life

This paper improves the estimation of the amounts of sewage flow, which is one of daily life data, in order to manage them efficiently. The amounts of flow of a typical day are tried to be adjusted to those of a non-regular day. A typical (non-regular, respectively) day is a non-rainy day having goo...

Full description

Bibliographic Details
Published in:International Journal of Networked and Distributed Computing (IJNDC)
Main Authors: Teruhisa Hochin, Hiroki Nomiya
Format: Article
Language:English
Published: Springer 2017-09-01
Subjects:
Online Access:https://www.atlantis-press.com/article/25885021.pdf
Description
Summary:This paper improves the estimation of the amounts of sewage flow, which is one of daily life data, in order to manage them efficiently. The amounts of flow of a typical day are tried to be adjusted to those of a non-regular day. A typical (non-regular, respectively) day is a non-rainy day having good data and no (a few) outliers. The values for the adjustment are tried to be estimated by using the multiple regression analysis. It is shown that the estimation can be improved, and these values can be estimated by using the temperature of that day, the amount of the rain fall of the previous day, and the day type, which distinguishes a weekday, Saturday, Sunday, and a national holiday. The estimation is tried to be used in estimating the data of a regular day. It is experimentally shown that the estimation works well.
ISSN:2211-7946