Smoothing 4: Simple exponential smoothing (SES)

Simple exponential smoothing is a popular data-driven method for forecasting series with no trend and no seasonality.
This video supports the textbook Practical Time Series Forecasting.
www.forecastingbook.com
www.galitshmueli.com

[@FlareSoul69]

Finally!!! This lecturer is highly technical, but does a superb job of explaining everything!!!

[@ahmedashry4123]

You are more than I could imagine , simple , clear , short and also digged deep.

[@cananakca1408]

Your English is so perfect, and your voice is so clear, I could listen to your training forever...The way you communicate the topic is superb. As a beginner in the topic, I understood right away.

[@xinyichen3554]

It's really a great series of videos explaining time series analysis!!

[@matthewhoag9892]

This is fantastic - thank you, Galit!

[@skrishnakumar00]

Thanks Galit for such a clear explanation ! The visuals - formulae, example and ur explanation went really well together and helped to understand the concept easily!!

[@rajjubhaiwala4508]

Very well explained, thank you Prof.

[@yousefhassan7365]

Thanks for this great effort. Appreciated!

[@davidkim4989]

how beautiful and easy explain. Thanks

[@Matheus-vc2or]

Thanks, Very helpful!

[@letussustain2933]

God of data science 🙏🏼

[@HazemAzim]

although back in 2016 , but super explanation. sharp and crystal !

[@thusoetsilemodise4498]

fantastic staff

[@Josephblayil]

Thanks, Very helpful

[@oleersoy6547]

You are awesome!!

[@chandrak.5447]

super job.. thanks..

[@jokosulistyo3144]

thank you, great video :)

[@xinpang9611]

This is a fantastic video for studying time series! I have one small question: at 6:18 the teacher explained that trailing MA with window W almost equals to SES with alpha = 2/(W+1). I did not quite understand this part. Could anyone please explain a bit on how the conclusion is derived? Thanks!

[@rodrigoavila5580]

Amazing video

[@arunborgohain7372]

Nice explanation ⚘

[@iiilllii140]

super useful, ty!

[@glennsugana4348]

So how exactly could you forecast the values beyond the 1 period immediately succeeding the last available historical data?

[@VideosEngracadosBR]

At 3:35 L(t) = ... should be L(t+1) = ... .Do you agree? (nice lecture by the way)

[@minniemouse5960]

By Lt and Lt-1, I assume this is the same as notation in your book that uses the Ft and Ft-1 for the forecast value?

[@PinkFloydTheDarkSide]

But you only gave an example for SES where you used the last forecast value, alpha, and the error. Where is the example on a weighted average for a series of past values?

[@virajvaidya1770]

What is exact meaning of level?

[@nato838]

You are a fucking treasure