I really appreciate people like you who take the time to make these videos. I have been struggling with this chapter for days now and here you're solving it in a way I understand better than the textbook described. Thank you!
Hello Mr Vakil, This is exactly what is needed for explaining 1) how to find whether your time series fits an auto-regressive pattern (i think this is not explained, but assumed that we have an AR model 2) if yes, what is the lag, e.g., AR(1), AR(2), ..and how to compute that 3) once you find out its AR(1) - "how to actually "forecast" the future time-value. Lucid, simple, hits home. I have been browsing, several vids, and this is exactly the kind of explanation i was looking for. Is there any video from you, which explain Moving Average prediction? Also can you extrapolate, and similarly lucidly explain, ACF and PACF? Other contributors, just dont talk about opening an excel and explaining with a simple series !!
Thanks, I instantly figured out what I was doing wrong. Looked at several videos and this was much more accurate and fit my situation better. Thanks again.
I've got a question: what if during the regression process both coefficients would be lower than 0.05 (like : first 5,8*10E and second 0.471). Do we reject te second (0,471)?
Great video! One important question: in the end of the video we see that an AR(1) is a good model for predicting value at T+1. Well. Now let's imagine I wanna see the PAC (partial auto correlation). Is it different in the computation? I mean how to subtract the n lags autoregression? Hope I made myself understood
I really appreciate people like you who take the time to make these videos. I have been struggling with this chapter for days now and here you're solving it in a way I understand better than the textbook described. Thank you!
Hello Mr Vakil, This is exactly what is needed for explaining 1) how to find whether your time series fits an auto-regressive pattern (i think this is not explained, but assumed that we have an AR model 2) if yes, what is the lag, e.g., AR(1), AR(2), ..and how to compute that 3) once you find out its AR(1) - "how to actually "forecast" the future time-value. Lucid, simple, hits home. I have been browsing, several vids, and this is exactly the kind of explanation i was looking for. Is there any video from you, which explain Moving Average prediction? Also can you extrapolate, and similarly lucidly explain, ACF and PACF? Other contributors, just dont talk about opening an excel and explaining with a simple series !!
very very helpful vedio. I learnt from it. with in 5 mintesthat how to run Autoregressive model. JAZAKLLAh.
Thanks, I instantly figured out what I was doing wrong. Looked at several videos and this was much more accurate and fit my situation better. Thanks again.
wow ur actually the goat! W mans fr
Very clear. Thanks!
Very well presented! Thank you very much.
Thank you for posting this!
Ragz B
, you just wrote my mind. I have the same experience. I also look for the same other videos.
I've got a question: what if during the regression process both coefficients would be lower than 0.05 (like : first 5,8*10E and second 0.471). Do we reject te second (0,471)?
How Alfa is taken as 0.5?
if i want to proceed with moving average , which order should i use ?
If alpha value does not given then what should I do ?
I did first order and my P-value is coming as 0.00. Did I do something wrong?
Great video! One important question: in the end of the video we see that an AR(1) is a good model for predicting value at T+1.
Well. Now let's imagine I wanna see the PAC (partial auto correlation). Is it different in the computation? I mean how to subtract the n lags autoregression? Hope I made myself understood
Thank you for this video it is very usful plz can you do for us an author vidéo for autocorrelation unsing excel
is it vector auto regression? or not?
great .
how to test autocorrelation after this?
You can run the Durbin-Watson test for the residuals.
how did you get the value of alpha?
Alpha is given. It is the probability of wrongly rejecting H0 = Prob (H0 is rejected given the fact that H0 is true)