ARIMA Models for Stock Price Prediction ❌ How to Choose the p, d, q Terms to Build ARIMA Model (1/2)

Microsoft Azure Certified:

Databricks Certified:

---

---

COURSERA SPECIALIZATIONS:

COURSES:

LEARN PYTHON:

LEARN SQL:

LEARN STATISTICS:

LEARN MACHINE LEARNING:

---

For business enquiries please connect with me on LinkedIn or book a call:

Disclaimer: I may earn a commission if you decide to use the links above. Thank you for supporting the channel!

#DecisionForest

[DecisionForest]

Hi there! If you want to stay up to date with the latest machine learning and deep learning tutorials subscribe below. Thank you for your support!

[sudharsan_balakrishnan]

This video really helped me to finsh my College project. The way you explained is just awesome. Thanks a lot

[vsvm]

Your video is extremely useful, I was suffering from having no idea of putting p, d, q parameters into the model. Thanks to your explanation that it makes clear about what these parameters are by applying differencing values and plotting out PACF to determine the parameters, compared to other lecture materials.

[danielb.]

Thank you for this. But you did not split the dataset into training and testing datasets which is the usual practice in time series analysis.

[prithvpradhan]

Can you explain how you chose q=3 from the graph. I could not get it.

[emresdance]

Amazing video, a hidden gem man, just subbed, thank you!

[tridunghuynh]

Thank you for your instruction. Could you explain how did you choose q value base on autocorrelation?

[robertaraujo]

I was taught that we can only use ACP y PACP when working with either AR or MA models separately, because when combining them this aproach to select p and q is not longer valid. The right aproach would be creating the ARIMA(1, 1, 1) model and analyzing the residual ACP instead of the two previously mentionated, am I right? or could you please explain me why are you using them this way?

[marcinfurga]

Very well explained and helpful but I do not understand one thing: what is the criteria of choosing p and q? Is it the value above the significance level laying the closest to zero or what?

[tillmannbrodbeck]

Thanks. I did not understand how you chose q though, is it the lowest value in the graph stading out?

[the_imonem]

@13:15 use "from statsmodels.tsa.arima.model import ARIMA" instead of from "statsmodels.tsa.arima_model import ARIMA" and "result = model.fit()"

[Espfs]

I kind of confused about this whole thing. The prediction line is always behind the price... It should tell us what's gonna happen before it happens... Not after its happens

[architsharma]

What about using Auto Arima which gets us the best order for p, d, q values for our ARIMA model? Will there be different answers if I choose your method to get each of p, d, q values separately?

[AquaMVP]

Is there way to store the value of p and q in a variable without having to analyze the chart like how we did for d?

[VitorHugoP]

Why would you make the data Stationary if you are using the original data to fit ARIMA?

[umesht]

The website and the Jupiter codes you have given is not working. Showing URL error

[himeshkoli]

you have no idea how much you have helped me, thanks a fuckin ton

[podunkman]

I understand nothing. There must be better, clear way to explain what is p and what is q.

[Lejik]

Hi, one question, how i check next prediction, it possible predict next day, or next few days, this is what i not understand in this model. All models which i did before, like linear regression, it was easy, i take features or feature split data for train and test, and after i add some data to predict it, here i not understand how to do it.

[nazmulshohan]

You're awesome teacher ! take love brother.