02417 Lecture 6 part B: Identifying order of ARIMA models

This is part of the course 02417 Time Series Analysis as it was given in the fall of 2017 and spring 2018.

The full playlist is here:

You can download the slides here:

The course is based on the book:

[brakit]

All ACF and PACF interpretation videos show very clean graphs that are easy to interpret, so its never clear how to interpret ones that are not so standard. This video explained so easily how to do it. I can't believe this is the only one I have come across in all my searching that does this.

[stojanovich]

Phenomenal video! I've spent half a semester on ARIMA in my business forecasting methods class for my Masters program and I didn't understand what was going on until I watched this video. You are a great teacher. Thank you for the content!

[AJ_]

The best explanation out on Youtube on this subject till date!

[pranavkumar]

I was really confused about interpreting the ACF, PACF plots. This video helps a lot. Thank you :)

[henryl]

thanks so much! straight to the point!

[egehanyorulmaz]

It helped! Thank you for giving examples from all the possible scenarios.

[dungtran-qmnt]

Thank you a lot. The video and your examples are clear and easy to understand :)

[ztac_dex]

good info for people who are crash coursing this

[drajabsingh]

Sir,
Thank you!
Very much
ARIMA Model Identification aspects and issues are expressed more clear.

[hawrezangana]

You are an amazing professor! Thank you!

[tomasharasta]

Damn, why is it always youtube that explains the topics better than my professors

[eliasbojago]

Graet explanation. I need the PPT/DOC too.

[veronicatrabacchin]

Thank you for explaining these concepts so clearly!!

[dradfulboss]

Thanks, nice video. Clear and to the point.

[TimelyTimeSeries]

Thank you so much for the video. The examples helped me understand the concept much better!
I have a question, though. In 07:53, there is one significant PACF, that's why you consider it as AR(1) process. However, that PACF is in lag 5. Why is it not AR(5)?

[toyosiojo]

what if the trend is not as easy as these? what if there's no exponential decay on both plots?

[sebastian]

Very useful information compress in just 13 minutes, thanks a lot! . Btw I lived 1 year in Odense back in 2018, if I'm not wrong you have Fyn accent right? I didn't expect to listen that accent searching tutorials for grid searching method haha. Greetings from Chile :)

[bhavinmoriya]

How could I choose seasonal order? I have daily data with period of about 365 days. Shall I take m = 365?

[badiaamakhlouf]

Thank you for the video and may I know if the PACF and ACF plots at the begining of the video are for order p and q or only for order 1?