So much for professors who like inflicting pain on student subjects by presenting these concepts in a convoluted way. Like if it's an esoteric black magic.. Step bey step crystal clear ! Thank yo and Bravo!
Why doesn't this video have more clicks?! Simply awesome. If my empirical research lecturer only were half as talented at explaining things as you are, I wouldn't need to be here right now. Thanks so much!
I am a graduate student in Civil Engineering. But I am fairly new to the subject area of time-series forecasting. This has been the best intro I found on RUclips. Not so tough on the mathematics but seamless in every aspect of the topic. Simple and clearly understandable. If I could give more than a thumbs up I would definitely give; recommended. Thanks!
After desperation from looking at the bunch of formulas in my study book, I'm so glad to have watched this video. Literally gave me AHA moments for 40 minutes :) thx
DIfferencing, I'll have to try that. Demand in electric utilities is MW or kW, it is a measure of power, not energy. When asked about demand, you are giving a instantaneous power consumption. A user will also provide a capacity or load factor which gives you kWh.
Upon watching them all, I think he numbered them not using the actual lecture number but session no. The missing classes were probably those he used to show an example or have them code something.
At 34:37 you said when the correlation for two time series is 0 it becomes statistically independent, I beg to differ because the correlation only gives you an idea about linear dependence, but not about the non-linear one.You cannot conclude that the two RV's are independent.
Thank you for posting this lecture. Your descriptions are nice and clear. I particularly liked the approach to auto-correlation. I did feel that the part covering seasonality needed more depth though. Is simply taking the mean for each month (which I assume you do in a step-wise manner rather than some moving average) a robust approach to considering seasonality?
Good evening my Professor, Please sir, if we have the Yt series. To study the stationarity of this series, we can do the following decomposition (or filtering): Yt=F(t)+Ut, such that F(t) is a continuous function according to the trend (linen, nonlinear). And if we find the series Ut it is stationary, it implies that Yt is stationary, and the opposite is right? B.w
Today I presented my master's thesis proposal on univariate forecasting (time series forecasting). I intend to combine different models such as Holt Winters, ARIMA, LSTM neural networks, random Forest, etc...The Synods told me that time series forecast are simple and less accurate than multivariate forecast... any advice to answer them? ?
If I am going to conduct a var model of variable 4 , and find that 2 variable is stationary and 2 variable is non stationary. Numeric Unit of those variable percent i.e. Interest rate, repo rate and deposit rate etc. What to do?
HELLO, anyone who can tell me that is their any restriction of sample size in time series data? like i have taken data from 2004 to 2017 so the total number of observations are 14. So in this situation time series method is applicable or not?
Sundas Memon Well yes the method is applicable but it will not give a lot of insight. With only 14 samples there is not a lot of information in the data. Furthermore, you must go back to the underlying proces that generated the data The guy does not grasp the fundamentals of digital signal processing
This is the best and most comprehensive intro to time series that I've found on youtube. Thanks so much for making this
Right
So much for professors who like inflicting pain on student subjects by presenting these concepts in a convoluted way. Like if it's an esoteric black magic.. Step bey step crystal clear ! Thank yo and Bravo!
Best time series lecture I've come across so far. The real life examples makes understanding the concept way easier.
Why doesn't this video have more clicks?! Simply awesome. If my empirical research lecturer only were half as talented at explaining things as you are, I wouldn't need to be here right now. Thanks so much!
I am a graduate student in Civil Engineering. But I am fairly new to the subject area of time-series forecasting. This has been the best intro I found on RUclips. Not so tough on the mathematics but seamless in every aspect of the topic. Simple and clearly understandable. If I could give more than a thumbs up I would definitely give; recommended. Thanks!
Time series analysis is of utmost importance in Finance and Economics
This is the best time series introduction courses I've had! Thank you very much!
After desperation from looking at the bunch of formulas in my study book, I'm so glad to have watched this video. Literally gave me AHA moments for 40 minutes :) thx
Very useful video, well explained and illustrated. Thank you very much for making it available to the wide audience :)
The most informative video on time series by far on youtube, thanks a lot!
Well done!! This was an incredibly well written and well structured explanation of a complex topic. Thanks for posting it!!
Would it work on counts? Stellar job. Just the right level of detail, perfect pace, and organized layout. You really take the listener by the hand.
Fabulous lecture. Everything is so clear! Thank you for making it available.
Excellent video! I'm glad I found your channel, now I have to watch your videos, they all look very interesting !
the best explanation of time series analysis ever! after looking for a good one for one month. thank you Jordan
BEST video on time series! Period!
DIfferencing, I'll have to try that.
Demand in electric utilities is MW or kW, it is a measure of power, not energy. When asked about demand, you are giving a instantaneous power consumption. A user will also provide a capacity or load factor which gives you kWh.
Thanks for sharing this information. Comprehensive concepts intro and easy to understand. Great for my job project at this moment.
Thank you for the most comprehensive lecture I've ever seen!
This is super helpful! Thank you so much! This is perfect for getting an overview for me, starting at the bare basics!
Just 12 min in and I agree with everyone else. This is the best lecture I have ever seen on time series. Thank you so, so much.
Dammit man, post the rest of your lectures!
Upon watching them all, I think he numbered them not using the actual lecture number but session no.
The missing classes were probably those he used to show an example or have them code something.
Thanks!! I've been through 7 minutes and I love it!! Thanks for this video..!! Will certainly share with my friends!
true
At 34:37 you said when the correlation for two time series is 0 it becomes statistically independent, I beg to differ because the correlation only gives you an idea about linear dependence, but not about the non-linear one.You cannot conclude that the two RV's are independent.
Amazing job with introducing time-series relative to regression!
Thank you, this is extremely useful and very accessible. A great introduction to the topic!
Also this link at the end! www.itl.nist.gov/div898/handbook/pmc/section4/pmc4.htm
Thank you, and by the way 11:49 there's a clear trend on a first graph indicates about Batman existence!
Thanks dude, finally understand some concepts!
Helped me understanding a lot more than in my lecture! Thank you very much for creating this great content
Ur my life saver in understanding this topic
Finally, autocorrelation explained simply! Thanks!
the best overview of time series concepts. taking it from a guy doing a statistics degree.
Incredible lecture. Thank you so much.
#amflearningbydoing #amflearning this awesome bro, thanks a lot
Excellent lecture Jordan
The best ever! Great job, well done. Thanks a lot. :)
This is very good intuitive intro!
best lecture I have seen this year. grattitude for your dedication
thank you so much, i needed this for my econometrics class big time.
@Jordan Kern Do you recommend any reading material to accompany your wonderful lectures?
Best lecture in time series so far
Very good explanation of TS, Jordan. Nice job.
Thank you for posting this lecture. Your descriptions are nice and clear. I particularly liked the approach to auto-correlation.
I did feel that the part covering seasonality needed more depth though. Is simply taking the mean for each month (which I assume you do in a step-wise manner rather than some moving average) a robust approach to considering seasonality?
Really informative video
Straightforward and easy to understand
Excellent lecture! Chapeau!
Normal distribution or g l m😊
If you have 3 points up in a row it doesn't mean that it has a memory. Saying that time series have a memory - it something which should be proved.
Very talented lecturer
Good evening my Professor,
Please sir, if we have the Yt series. To study the stationarity of
this series, we can do the following decomposition (or filtering):
Yt=F(t)+Ut, such that F(t) is a continuous function according to the
trend (linen, nonlinear). And if we find the series Ut it is stationary,
it implies that Yt is stationary, and the opposite is right?
B.w
great work
Today I presented my master's thesis proposal on univariate forecasting (time series forecasting). I intend to combine different models such as Holt Winters, ARIMA, LSTM neural networks, random Forest, etc...The Synods told me that time series forecast are simple and less accurate than multivariate forecast... any advice to answer them?
?
Thank you for your service 💞
If I am going to conduct a var model of variable 4 , and find that 2 variable is stationary and 2 variable is non stationary. Numeric Unit of those variable percent i.e. Interest rate, repo rate and deposit rate etc. What to do?
Hi Jordan - how do you produce the last plot you showed with relative variance vs. frequency?
Hi Mr jordan Sorry, is there a pdf file for this topic that I need?
well explained.
It was amazing course
Fantastic!!!
Excellent. Thank you
great video!!! Can u share slides, it will be nice!!!!!
Great work!
Great!
This is statistical poetry.
Your teaching material is awesome. Could you please share the file?
wonderful, its now easier!
phenomenal sir
At 1.5x, the guy becomes John Krasinski
Is Smoothing done on the 'White noise' which was obtained after removing the signal data?
Thanks 😊
So, it can be this simpler? Cheers!
Thanks, this lecture was very useful for studying econometric. I'm spanish, by the way.
You’re amazing
HELLO, anyone who can tell me that is their any restriction of sample size in time series data?
like i have taken data from 2004 to 2017 so the total number of observations are 14. So in this situation time series method is applicable or not?
Sundas Memon
Well yes the method is applicable but it will not give a lot of insight. With only 14 samples there is not a lot of information in the data.
Furthermore, you must go back to the underlying proces that generated the data
The guy does not grasp the fundamentals of digital signal processing
Where is lecture 1 I would like to start
with the background noise, it sounds like a recording made in 1950s :) . Nevertheless very informative and interesting presentation, thanks.
13:09
Noise: 🚓 🚨
Why are there so many lectures missing? I enjoyed listening to these at 1.5X speed.
lol, same, was listening at 1.5x speed.
👍
Histrogram😊
Garcia Dorothy Rodriguez Brian Brown Larry
Thanks!!!!
Mani pavi call Panna vaa attention Panna matiya true caler natrajen varum
12:55 this man risking his life for our stats education
I still don’t understand
Pretty good but I wish he did not explain time series and regression to be exclusive topics. Regression is merely a scientific objective.
Can I have your email Sir?
wonderful, its now easier!
wonderful, its now easier!