ARIMA Model Explained | Time Series Forecasting

MY STUFF:

CONNECT WITH ME:

TIMESTAMPS:
0:00 Intro
1:05 What is ARIMA
5:35 Requirements for ARIMA
6:40 Fitting ARIMA models
11:10 ARIMA in Python
15:25 Recap
15:54 Outro

DISCLAIMERS & DISCLOSURES:
This content is for educational and entertainment purposes and does not constitute professional advice. Opinions are my own. Some links are affiliate links that support the channel at no extra cost to you.

[danielfonseca]

Man you are the best... Really struggled to find good content explaining time series concepts and applications... Most sources fall short on one of those, but your material is perfect. Many thanks 🙏

[AzureCz]

I can't be the only one who hears "Hi Amigo" everytime on the videos hahaha (high quality classes btw <3 )

[gwengaw]

Great video!!!! I wish I found this first!!

[scondon]

Hi Egor, Thank you. It sounds like some human intervention/judgement is required to select the values of p, d, and q (i think those were the names). Is there a way to automate that selection so you can compute a forecast without human intervention? I think you mentioned a method that evaluates all of the possibilities and suggests the one with the best fit. Can you elaborate or refer me to another video?

[cc]

Nice video which is easy to catch up with. Keep it up.

[paxerk]

Do you have to feed the data that was differenced to the ARIMA model or the original? I remember you saying that some models apply this differencing automatically, but I am not quite sure about this one.

Also, how come you are using the Passengers_Boxcox and not Passengers_Diff when you are feeding it into the ARIMA model?

[donharrold]

I'm guessing that forecasting far beyond the original data set is somewhat risky, i.e. if you have 4 or 5 years of data then it may be sensible to forecast for 6-12 months beyond the original data set???

[victorfigueroa]

Thanks for this video it really helped me out.

[DavidEsparzaAlba]

Great videos, BTW it sounds like you say "Hi Amigo" :P