What is Time Series Decomposition
HTML-код
- Опубликовано: 1 окт 2024
- The second video in the series on time series. This video covers the topic of exploring your time series data - time series decomposition. It talks about trend, season, and error as well as some extra information on LOESS regression for calculating the decomposition.
by sticking to the 5 minutes constraint, you proved that you are time serious about time series (my dad joke attempt)
Your series are a monument of Intuitive Understanding of concepts.. This is what practicioners need!
Cool video! Please don't stop with the series! It's very helpful!! Thanks!
Thank you! Glad you like them!
This is pure gold - have studied risk, trading, statistics, this is next level due to both simplicity, visualization and awareness/understanding of the discipline and concepts communicated. Thank you...and it is amazing
Thanks very much! Is LOESS regression the same as Seasonal Trend Lowess Model or is it a bit different? And for each individual observation that you find the linear regression line from, do you need to create a new window so that that observation point is in the center, or is that not necessary ?
It is the same idea at the STL model (Seasonal Trend LOESS)! The seasonal piece is found by LOESS by looking at the specific season values (for example, all March values). Once this seasonal component is estimated it is removed (think seasonal adjusting like I described in the video) and then we do LOESS to on the remaining piece! Hope this helps!
is it sheldon cooper talking....?
amazingly explained!! Thanks
finally understood that **** LOESS regression. Thank you so much
Wow, great work!
That was perfect.. I was thinking, "What the heck is Loess regression?".. then, boom.. lol. Thanks.
Glad it was helpful!
best teacher
Great refreshers. I'm subscribing.
I was reading that LOESS regression is best for additive decomposition, but based on this video, it seems like both additive and multiplicative can use LOESS. Is there any limitation on when we won't use LOESS?
You are correct that it is better for additive, but still available for multiplicative. The real difference between the multiplicative and additive is how you remove the effects from the model. Multiplicative, you divide the effects away as compared to additive you subtract!
Very nice video. keep up the good work. Thanks!
This series so good! Clear explanation and examples! I do wish you keep doing this for other topics too 🎉
That's the plan! Full time job gets in the way :-)
Highly precise information in just 5 mins. I wanna see you gain more popularity!
ah thank you, I've been lost with my online class
pls upload more often
Definitely will try to! This is only fun on the side of a full time job :-)
Where have you been all this while I've laboured in vain to easily comprehend time series? Please I implore for more videos; I'm following and waiting for notifications of new videos
Thank you Victor! The time series section is full for now, but I have other series that I am publishing!
@@AricLaBarr please you have my permission to tag me when you publish any new video. Thanks.
Keep up the good work!
Very helpful sir
Brilliant.
great job.
i really like your videos. I have seen a lot of time series videos, you are the best.
Glad you like them! Thank you!
appreciate the video , well explained
Great and concise explanation, loving it
nice
very good
You, sir, are awesome
Excellent!
Glad you liked it!