How to implement exponential smoothing for data without a trend/seasonality?

15 views (last 30 days)
Hi everyone,
My data is the form of AIS and I'm only focusing on Speed (SOG - Speed Over Ground) and Course (COG - Course Over Ground) parameters as a time series. I'm attaching a csv file that contains the time stamp (yyyymmddhhss) and SOG values for reference.
What I want to do is implement exponential smoothing to predict the next values. For the reference: the math behind this is as follows.
My question is:
Which exponential method will suit this the best? (I followed a bunch of articles and most of them say that simple exponential smoothing is ideal for data without trends/seasonality. But the results I got from that is not accurate and I would like to know if I can implement triple exponential method)
Is there a MATLAB method to implement this? (I found Holt Winters tool here, but it does not seem promising)
Also, is there a way to manually implement the above equations using MatLab? I'm not sure if I know how to carry out the first iteration :(
I'd appreciate any tip/recommendation. TIA!

Answers (0)

Categories

Find more on Mathematics and Optimization in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!