nice cours , thanks

Course is very good.

Excellent course

learned a lot!

Best Course.

nice course

Fantastic

VERY GOOD

next onE

Great

Good

nice

A good course for learning data analysis skills. My one critique is that model evaluation section was possibly a little overly complicated for non-academic professionals(for example I found lasso and ridge modeling very difficult to understand). Besides that it was a very interesting course.

The only problem for me was understanding that lab practice problems are in Jupyter Notebook, instead of the R script! Strange but dealt with it!

this course is not for the week, its not challenging but you have to litle dictated...

This course was more theoretical than practical

A few issues but fixable.

Honestly, I was a little underwhelmed by the way this course was delivered. There were a number of new and complex topics that were introduced but were very poorly discussed, including tidy models using recipes, and regularization using lasso and regression. Some of these things were quite new to me, and it would have helped to get specific practice. Alternatively, it seemed the course tried to cover too many things, and things like the recipes could have been dropped. I also did not understand why I ended up spending so much time setting up a Watson account, when my time was limited on it, and I could have done the project in the Jupyter notebook anyway.

I actually really liked the course.....until the final project!!!!!!!!!!!!!!! I spent more time dealing with the IBM WatsonX and GitHub b.s. than that actual final project and I never found out why. In the end all I had to do was submit screen shots of the code, which I could have done from Jupyter. The real issue is that in WatsonX there was an issue with the tidymodels library so I could not run most of the code. Now why would you have us work in WatsonX, tell us to load tidyverse and it wont use 3/4 of the functions. Anyway, if you either got rid of WatsonX part or had WatsonX have the tidyverse lobrary fully functional then this was outstanding.

It might help students if a portion of this course were required prior to the SQL etc courses.

The AI grading was incorrect. E.g. 0 points were given, although task was correctly done

las bases del trabajo final no funcionan

The first four modules are pretty good, albeit fast-paced. Module 5 presents a lot of information too quickly; it would be better split across multiple modules and more labs. The AI grader for the final project seems overly harsh/buggy and it was challenging to figure out what needed to be included in the screenshot to get credit.

lots of technical issues! :,( for example, IBM db2 Lite is no longer a thing, many code chunks include deprecated code, Watson Studio did not have a place for the redeem code without using credit card. Also, some of the labs had paragraphs out of order, was very confusing.

This course is a mess. I worked a lot with R before - mostly self taught. So my motivation of taking this course was that I wanted to get an official certificate proving my knowledge. Generally speaking, while the way they code here in this course may be fine for little projects, it will cause you to experience issues in bigger ones, because it is highly inefficient - mainly through unnecessarily defining new data frames in each and every step. When coding this way, you will get very high run times if you work with even slightly larger data sets, and frankly, chances are you may be not able to grasp your own code at some point when working more complex projects. Secondly, you will at several points encounter stuff that does not work properly - e.g. some codes in the labs, the link to the data set for the assingment. This was quite frustrating, honestly. Overall, I spend more time setting up their IBM-own Watson studio than on any practice lab. Do people at IBM think, that making us use Watson studio for their courses will lead to people actually deciding to pay for that service in the long run? Concerning data analysis projects - after this brief look - I would say RStudio offers by far a better environment. Third, the statistical concepts are poorly explained. If you already have good knowledge of those topics and only search for a way to get certified you might be pleased by the pace. But good luck if you don't. In some ways, one might think it is quite arrogant to think that one can explain complex topics like overfitting/underfitting/L1- and L2-penalties and their implications, which are handled in several lectures in real courses, by using 2-3 videos of less than 10 minutes. I did not understand, why this was even discussed in this course, especially after seeing how superficially regression models were presented. I am not sure if the instructors even have a good understanding of the mathematics in place here. Overall this course fulfilled my personal objective, but I would highly recommend not enrolling in this course if your knowledge on the topics discussed in this course is not already sufficient.