IMPROVED ANALYSIS FOR TIME SERIES WITH MISSING DATA

Authors

  • Mukta Agarwal Assistant Professor, Sabarmati University, Ahmedabad , Gujarat, India,

Keywords:

Time series prediction, missing data, ensemble prediction

Abstract

Particular range examination (SSA) is an incredible method for time arrangement investigation. In light of the property that the first run through arrangement can be imitated from its essential parts, this commitment builds up an improved SSA (ISSA) for preparing the fragmented time arrangement and the adjusted SSA (SSAM) of Schoellhamer (2001) is its unique case. The methodology is assessed with the manufactured and genuine inadequate time arrangement information of suspended-residue fixation from San Francisco Bay. The outcome from the manufactured time arrangement with missing information shows that the overall mistakes of the important segments recreated by ISSA are a lot more modest than those remade by SSAM. Additionally, when the level of the missing information throughout the entire time arrangement arrives at 60 %, the upgrades of relative blunders are up to 19.64, 41.34, 23.27 and 50.30 % for the initial four head parts, individually. Both the mean supreme mistake and mean root mean squared blunder of the recreated time arrangement by ISSA are likewise more modest than those by SSAM. The particular enhancements are 34.45 and 33.91 % when the missing information represents 60 %. The outcomes from genuine fragmented time arrangement additionally show that the standard deviation (SD) inferred by ISSA is 12.27 mg L?1 , more modest than the 13.48 mg L?1 determined by SSAM.

Downloads

Published

2020-10-31

How to Cite

Mukta Agarwal. (2020). IMPROVED ANALYSIS FOR TIME SERIES WITH MISSING DATA. European Journal of Research Development and Sustainability, 1(2), 25-28. Retrieved from https://scholarzest.com/index.php/ejrds/article/view/43

Issue

Section

Articles