This got a little too technical a little too quickly. It doesn't tie any of the abstract operations to real-world interpretations, so I got a little lost in the rapid series of transformations. The practical examples were interesting and helpful, but they came a little too late in the course for me to be able to juggle all of the concepts successfully. It's almost like having someone read a textbook to you.

The material was good, although sometimes it was hard to see the "practical" side (now that I've modelled the time series what do I do with that?). The course also needs to be updated as some of the links no longer work. It would also be helpful if the two professors coordinated their efforts better - flipping back and forth between two instructional approaches was frustrating.

Could've been more technical, with introduction to new machine learning techniques like RNN.

This is the first, and last, Coursera course i will do

This is a good class. Well composed, and covers the material in a reasonable manner. Overall the best points of the class were Professor Thistleton's readings, which were very well put together and did a nice job of developing concepts. If you push yourself to follow along with them you'll develop a very sound conceptual basis about the material. Likewise, Prof. Sadigov's exercises with his notebooks were also useful, but his notes not so much.

If the class has some weak spots it's that, like a lot of classes on Coursera, the amount of time you have to spend to pass the class can be quite small if you just want to cruise and finish the course. Also, even though it's called "Practical Time Series Analysis", the majority of the material is quite conceptual. This class is a lightweight attempt to expose you to the material you'd cover if you studied it in college, but it does not just dig in and show you how to start hacking away at models. You'll need to practice a lot more on your own to develop yourself as a practitioner.

It was the first time I deal theoritacally and practically with Time Series. It is a perfect course for a beginer.

In my opinion this course need some prerequisites in Calculus but even if one doesn't have he can complete ther course. In addition this will be a stimulus to build the adequeate mathematical backround.

Finally I would like to refer that just completing the course doesn't mean that you have aquire the pertinent knowledge. On the contrary you have to do a lot of on the job practice with reference the material of this course.

The real data for practice is the date every one finds in his occupation e.g sales , production and the like.

To the Proffesors of this you I would to refer that in the update of this course it would be very good to include a Week with Regression with time series and some theorhy and practice of detrending.

Thank Indeed.

Very well explained! Both instructors are very competent and are complementary in their way of explaining and simplifying complex concepts. It took me five days to complete this course and I really had fun going through the theoretical details behind widely used techniques such as ARIMA and SARIMA, as well as understanding the applications of the techniques. The course is well structured and easy to follow. There is just enough theory to understand the basic concepts, applications being the main focus of this course. Congratulations to the instructors for this great course and thanks a lot for putting all the efforts to make it accessible to people that have not worked with time series before!

This is a great course that provides strong introduction to time series analysis and forecasting. I have benefited a lot from it as I took it to advance my career in data science. I have found the mathematical formulations in time series analysis very useful. I have also found the forecasting sections equally useful. All quizzes and in lecture questions were very helpful. The R coding practices are certainly helpful in learning the corresponding R libraries; they also provide template code that is useful for writing custom code for analysis.Many thanks to the instructors!

I enjoyed and highly recommend the course! Both instructors explained the main concepts of time series analysis clearly. The practical aspects involved using R on various data sets.

The multiple worked examples were very useful in clarifying the concepts. The exercises and quizzes were generally direct applications of the examples. and thus very useful in helping to reinforcing the concepts learned. Clear explanations were even given for why certain choices were correct/wrong, which is not often the case for other coursera courses.

I hope for a more advanced course soon!

The course delivers on its promise: it is indeed practical. I had started several times courses on time series analsis but hardly finished one. In this case, the increments in model complexity are fairly gradual. There is a good balance between the required theoretical explanations and the application examples. I appreciated having the theoretical parts in writing before hearing them in a video. I also liked the fact that algorithms were first implemented in base R rather than directly using an off the sheves solution from stats or another packages.

I enjoyed the course, especially the theoretical part.Also I would wish there would more course, on Time Series Analysis at Coursera. Currently there is only one such course.In this course, I wish there would be more reference to the literature. Some points as determination of AR & MA order by looking at ACF & PACF plots is not clear enough to me. As I understand there is some rule of thumbs but deeper explanations are missing to me (i hope, that they exists).Anyway in my opinion is the best course in Time Series Analysis, that I ever had.

The first three weeks it is hard to understand as the course content was not properly organized. some chapters and quizzes are jumbled without order. It has a lot of theory as well. But then after understanding the basics, the theoretical concepts, it is easy to follow. It gave very confident and in fact already started applying in my real world time series problems to model and forecast for future time period. Great course and would recommend to friends who are serious to learn about practical time series analysis.

Despite some issues re data availability the course is excellent, well-structured and explained. It details the mathematics of time-series analysis and builds R programming skills. It uses R code to explain the key concepts which adds to the practicality of the course and to the understanding of the key concepts. The R code has changed since the course was developed and needs to be updated for R4 versions. The Discussion threads have the answer for these sometimes thorny issues around current R code versions.

The course structure is well organized from basic statistics to more advanced materials. I used to hate reading but in this course I found the reading materials quite pleasant and interesting. Only light coding involved so I guess people without coding experience would find it friendly as well. Both theoretical and applied aspects were discussed in details, and I got to know many valuable sources of finding interesting time series datasets. In summary, a really great course one must take a try!

Decent course with a right balance between math, coding and high level explanation. AR, MA and ARMA (ARIMA) models are very well explained. I am not a big fan of R (even after this course) but it seems its time series analysis libraries and datasets are comprehensive. The best part of the course is the in-course coding examples and tasks. They really help you get hands-on into analysing various time series objects. A little more emphasis should have been made on forecasting.

Time Series Analysis can take effort to learn. This course was well structured and I enjoyed the learning experience. It requires intermediate math and statistics background for grasping the intuition behind the theoretical aspects but it also has a balanced approach by introducing practical analytics functions using ‘tsdl’ , ‘astsa’ and ‘forecast’ libraries in R studio so one can become immediately productive and have some practical hands-on experience upon completion.

A good lesson to cover what under the hood of additive models like ARIMA and Holt-Winters Exponential Smoothing etc. It gives you the idea of what to approach time series, mathematical internals and some basic proofs. It helps you a lot to decompose a time serie in terms of seasonality, trend, autoregressive and moving average process components after making some statistical tests. It helped me to use it for my job to forecast some time series in a more accurate way.

Very well structured course and comprehensive study material available with the modules. Despite these, I found some gaps in practical support, but I did my own because I am familiar with the R software environment, overall it was a great experience. I request the teachers and team of this course to kindly make its second version module for advanced time series models (eg, ARCH, GARCH, ARDL etc.). I would like to join that too. Thanks!

First I started out reading Intro to Time Series and Forecasting, the book suggested by everyone. But, I could not understand the math because it was too tough. I did not lose hope. I completed this course because sometimes you need to get an overview of what needs to be done and then if you dive into the math of it, it will be easy. Much recommended course for the beginning of time series and forecasting techniques. 5 stars! Thank you

The course is very good for an introduction to time series. Few drawbacks are listed.

1. The theory behind double and triple exponential forecast are not given in the materials.

2. Some datasets are not available anymore in the Datamarket website anymore, needed to be fixed.

3. The forecast module in week 6 is kind of wired. Need more lecture to talk about what's the difference between smooth forecast and SARIMA model prediction.

My first MOOC to finish ever! This was an amazing course. Although some have complained about there being too much math and not being practical enough, I disagree. With a bit of patience the math starts to make sense, which makes the practical part more palpable. Finally, both of the instructors Tural and William where on point everytime explaining all the concepts and generally appeared like smart guys. Thanks a lot!

The course is exceptionally good. Just one thing that does not add up in the end; after a dense focus on SARIMA, we have just started to do forecasting with Holt Winters methodology. A relational information, or an explanation on the relation between them would be appreciated. One may be superior to another in some conditions or they are complete substitutes vice versa. In any case, the course was more that helpful.

The course structure is really nice and focused on hand's on application of Time series analysis. I was able to understand the maths also quite well, thanks to the Tutor for such a simplified explanation. I would look forward to see some more advance Time Series Courses like this.

I would highly recommend this course to all the active learners willing to learn Time Series Analysis.

Excellent course. The whole topic is broken into bits and pieces and in a well structured form. Makes it easier to understand each concept. Apart from quizzes, few assignment problems where from the given data-set you need to come up with which technique to be used whether SARIMA or Exponential forecasting and then final forecast numbers. This will make it a complete course.

It's really good. I'm a master of applied statistic student and haven't taken time series. It helped me a lot, the math can be decently challenging but was rewarding when I did it. The vocabulary are very obtuse though. Trend, stationarity, etc.. is lost in the details. Process order and lag relationship was somewhat lost in the detail. Overall I learned a lot thank you!