Master Time Series Analysis and Forecasting with Python 2026 (Udemy.com)

Until now (Section 4) its okay. But when the author talks about rolling average, people with no background of analyst/analytical thinking or statistics, might not understand, so it will be nicer for a 1 min explanation , using may be a excel with 100 days of data, and show whats rolling 7 days and rolling 30 day mean

I’m progressing at my own pace, but I’m thoroughly enjoying both the depth and the structure of this course. The content is thoughtfully designed, and the clarity in explanations really stands out. It’s evident that a significant amount of effort has gone into building this, and it truly deserves recognition.

It was one of the best courses I’ve ever taken. The instructor is genuinely knowledgeable and deeply experienced in the field, with a rare talent for drawing you into the subject. Studying stops feeling like a chore and becomes a pleasure. Before you even realize it, you have learned, and learned well. To appeal to a broader audience, the course keeps theoretical explanations on the lighter side. Personally, I would have liked a bit more depth in the theory. Even so, I can confidently say this is one of the best courses I have taken on Udemy. Many thanks to the instructor; this is a must-take course for anyone looking to build solid knowledge in the field.

Honestly one of the best courses on Udemy. As a junior data scientist who has no experience in handling time-series data, this course basically covers everything I need to know to start working on project on my own. The content is comprehensive, all the mainstream models are covered. I especially love the teaching style, because the instructor records every step of him doing the real coding, making mistakes, and correcting them. I had a lot of fun coding along.

Good summary level detail on applied time series using deep learning/machine learning. Unfortunately, most of the applied lessons are spent waiting for the instructor to look up needed commands or debuging errors that clearly were not prepared before class. This comprises much of the 33 hours of instruction. Also the remedial python tutorials are required to be completed in order to get the class certificate. Since the classes are taught assuming a knowledge of python, these should have been optional. Not why I took this class, and not very well structure.

This course is a bit of try, fail and correct all the time. It's annoying since I want a lecture that is right from the start. I know he says that he wants to share the pain, but I think it should be optional. Next I am not convinced that he understand the model for neural networks since the "b" is just called a parameter and a constant. It's the bias. Why not use that word. Why not use 10 min to explain how this equation came to be since it's the foundation for the rest? He does not explain it probably. If you want a detailed explanation of how the bias work, how the w (weights) come into play you should look at the foundation course by "Loek van den Ouweland" called "Neural Networks In Python From Scratch. Build step by step!" If more editing had been put into the course to cut out the failures it would have been super duper great. When I purchase a course I want as little BS as possible. Every lesson ends with "until next time have fun". It's not needed to say each time. Maybe after a section is done, but that's a minor issue.

Great course. I really liked the way you explained everything and provided a step-by-step walkthrough. I also appreciate that you shared all the Jupyter notebooks it made things much easier, especially for me, as I was quickly learning concepts relevant to my current job. By the way, it would be great if you could provide a simplified cheat sheet or document comparing all the models we studied. This would be especially useful for those looking to adopt new models for work. The cheat sheet could be structured as a table with models as columns (e.g., ARIMA, SARIMA, SARIMAX, TFT, N-BEATS, etc.), along with their pros, cons, and use cases essentially a summary of what you’ve been commenting on at the end of each course section in one document. Additionally, sharing the presentation files you used would also be helpful. Thank you once again :)

Diogo's explanations are outstanding—clear, concise, and easy to follow. What truly sets this course apart is its hands-on, practical approach. I often struggle to complete online courses, but this one kept me motivated and invested throughout. The inclusion of unedited mistakes and problem-solving moments was particularly refreshing. It made me feel less alone in making silly errors and taught me how to troubleshoot effectively. Another standout feature is the instructor’s responsiveness. Whenever I faced challenges, I could post my questions, and Diogo provided prompt and helpful answers. I even sought guidance on a project related to my work, and the advice I received was invaluable in steering me in the right direction. Overall, this course is a fantastic blend of theory, practice, and support. I highly recommend it to anyone looking to master time series forecasting in a fun and engaging way!

Many thanks, Diogo and Udemy for a great course! This will hopefully give people like me a strong basic foundation to advance our skills in forecasting through Python. I have recommended this course to a couple of people and will continue to do so.

The course was filled with various forecasting models and showed how to use them! The only suggestion would be to add more depth on the parameters/features that a forecaster might focus on to tweak the model. Fortunately, he dives into that - just wanted a bit more depth. This course was amazing! Thank you!