摘要
In order to extract deterministic component from trend nonstationary time series, regression analysis by Akaike Information Criterion (AIC) for segment size and mean size of cocoon filament is introduced, and deterministic component is extracted from size series of cocoon filament by analysis result. Experiments of simulating deterministic components on 9 cocoon categories are carried out, and experimental result is analyzed. Through analysis and experiment, it is known that selecting the order and coefficients of regression equation by AIC is beneficial to accurately describe the relation between segment value and mean value. This study is also useful for pretreatment of nonstationary time series.
In order to extract deterministic component from trend nonstationary time series, regression analysis by Akaike Information Criterion （AIC） for segment size and mean size of cocoon filament is introduced, and deterministic component is extracted from size series of cocoon filament by analysis result. Experiments of simulating deterministic components on 9 cocoon categories are carried out, and experimental result is analyzed. Through analysis and experiment, it is known that selecting the order and coefficients of regression equation by AIC is beneficial to accurately describe the relation between segment value and mean value. This study is also useful for pretreatment of nonstationary time series.
基金
the Ministry of Education,Science ,Sports and Culture ,Japan, Grant-in-Aid for Scientific Research (B) ,2005,(No.17300228)
关键词
蚕丝
尺寸
回归分析
化学成分
size series of cocoon filament
regression analysis
deterministic component
AIC
