Despite a long history of time series analysis/prediction, theoretically few is known on how to predict the maxima better. To predict the maxima of a flow more accurately, we propose to use its local cross sections or plates the flow passes through. First, we provide a theoretical underpinning for the observability using local cross sections. Second, we show that we can improve short-term prediction of local maxima by employing a generalized prediction error, which weighs more for the larger values. The proposed approach is demonstrated by rainfalls, where heavier rains may cause casualties.
Skip Nav Destination
Research Article| June 01 2022
Improving time series prediction accuracy for the maxima of a flow by reconstructions using local cross sections
Yoshito Hirata ;
Yoshito Hirata, Masanori Shiro; Improving time series prediction accuracy for the maxima of a flow by reconstructions using local cross sections. Chaos 1 June 2022; 32 (6): 063103. https://doi.org/10.1063/5.0092433
Download citation file:
Don't already have an account? Register
You could not be signed in. Please check your credentials and make sure you have an active account and try again.
Could not validate captcha. Please try again.