複雑な要因が絡んだ時系列データの予測に関する研究 : 予測アプリケーションソフトの開発

概要

論文の詳細を見る
The prediction method of complicated time-series data has not been established yet. The purpose of this study is to examine whether two methods are suitable to predict time-series data, and to develop the prediction software that used for those method. The method of least squares, the method of analyzing past data were examined. The result demonstrated that the method of least squares should predict tendency and the method of analyzing past data should predicting extreme value point for predict of time-series data near future. It is concluded that complicated time-series data should be predicted by three methods.
函館工業高等専門学校の論文