Learner Reviews & Feedback for Exploratory Data Analysis by Johns Hopkins University

My congratulations to all those people who worked to create this course although I have to pick up something I've found a bit annoying:

1- there were two video where the audio were nearly unintelligible

2- I would link the link proposed by the video to be possible to be clicked

3- Some exposition imperfection (even if they make these video more "real and human")

4- Since quiz are not so difficult to be evaluated automatically I found it a bit annoying to notice them locked by not-purchsing, even if I understand there have to be something which would make the customer to purchase.

I've found the swirl experience great although a bit annoying sometimes but I've no clue on how to possibly improve it so.

Keep up with this great work!

Bye

You'll learn some cool things like K-means clustering and creating dendrograms, as well as dimension-reduction techniques. The assignments are very easy if you have basic familiarity with R's base plotting system and the "ggplot2" package. I will say I'm very happy with this course in the overviews of R's major plotting systems (though no "ggvis" package), as well as working with color palettes. However, I wish there was more hands-on or peer-graded practice with K-means, heatmaps, dendrograms, and dimension reduction techniques like Singular Value Decomposition (SVD). If these are new to you (they were to me!), you'll certainly walk away from the course more knowledgeable.

Once again the teacher was really knowledgeable and engaging. The content was really helpful for my career. The part about clustering was challenging but still manageable. The pacing was good, not too slow (so not boring) but also not too fast (so still easy to understand). The case studies, especially the one about activity measured by smartphones, was one of the best parts of the course.

I didn't particularly enjoy some of the swirl practices. I found some of them to be very very similar (if not the same) as the examples in the lectures, so I only enjoyed the few where there was some new content.

Overall I really enjoyed the course and I would recommend it :)

A good start for data analysis, this course covers the basics of plotting with the three most common packages (base R, lattice, and ggplot2).I liked the assignment which difficulty is nicely measured (it is not just applying the videos concept, you have to look around the web to find tools and documentation about what functions to use).On a less positive aspects:- I am not sure this course was the best place to introduce kmean and PCA sections...- a lot of content is outdated (wrong links, old R command parameters, ...), look likes a quick freshup update would not do harm given the number of people that keeps registering...

It would be of the best interest to all that the content of the course be reviewed. Seeing references to data from 2012-2015 gives the idea that there's been no recent content review. Although not being the same as taking the full course at the university, this is still a paid training and a certain level of accuracy is expected.

Another note goes to the forums which should be cleansed or handled differently. It's not very helpful to check a forum to see that most of the threads are requiring reviews to the assignments, some from years back.

I have taken the course earlier, so am somewhat familiar with the layout and the materials. Overall it is a very good course and covers a wide range of subject matters. Roger has done a very good job explaining the concepts. I certainly would recommend this course to all who's interested in the subject.

I realize that there's limitation on the time people suppose to spend each week, however, I would like to suggest adding homework, in addition to quizzes.

Several video clips have some audio issues, not sure if that's fixable.

I like the case study part which provides you an overview of the practical applications of the skills learned in this course. Learned to use the different types of plotting systems in R which got me to use my 'hacking' skills to experiment and I find it fun. The least I like about the course is the clustering part, mainly because the topic is too advanced for my current level. But I am interested to learn more about it in the future. The swirl lessons were also really helpful to strengthen my understanding of the concepts.

A great course to begin with Exploratory Data Analysis. It teaches you how to analyse data and generate visual reports. However, to actually become efficient at Data Visualization one needs to dig deep and make use of other resources apart from this course. Also K means clustering and other types are explained well in this course but it would have been useful if there were exercises to help implement it in some real problem. Overall this course leaves you confident and enthusiastic about Data Visualization.

The pacing of this course was somewhat better than the ones that came before it. I felt that the depth of information covered and the questions asked in the projects and quizzes were a better match than previously. I still let myself take an extra two weeks to complete the final project because I was still learning and playing around with the plots and selection of data, but that was because I wanted to, not because I had to.

The course did a good overview of the different plotting systems in R, but it rushed through clustering. I had to watch the videos of k-means and hierarchical clustering at least 3 times to sort of understand it. The matrix concepts went completely over my head. Otherwise, the projects were very interesting, and I would highly recommend this course to other people.

Enjoyed this course a lot. This course allowed me to experiment with and practice various plotting techniques while analyzing the data in the initial stages. SVD and PCA were totally new concepts to me. It would have been better to see some real examples from the field with interpretations instead of understanding these concepts using random numbers' examples.

Course gives thorough introduction to basic tools for exploratory data analysis, including visualisation, PCA and clustering. Good mix of lectures, practical in swirl and programming assignment. Swirl practice are mostly a repetition of the examples in the presentations, which is a bit of a pity...and I missed a programming assignment on cluster analysis/PCA

Very good. Great videos but perhaps the most learning was obtained through seing different apparoches taken during the peer review. The course could be even better if more smaller peer reviewed tasks where to be completed where extra points where rewarded for not just displaying correct data, but also visualising it more efficiently.

I'm glad for completing this course, it added a value for me.

I wish the videos about (SVD and PCA) in week 3 was more clear but it was difficult for understand and i feel lost , I think you need to update this videos to have more a satisfied materials.

Thanks for your effort and for what i have learned for this course

Week 3 content is difficult to understand without background knowledge in clustering and component analysis. Hope the instructor can provide some materials or web links for cluster and component analysis at the beginning of Week 3. Other weeks' contents are good and helpful!

Great practical course on exploring big datasets in R. The main part, plotting, is very clearly and thoroughly explained and framed. Only 'single value decomposition' and 'principal components analysis' was somewhat hard te grab and need a lot of extra research and study.

I really enjoyed this course. I was a good reminder of what analysts need to do when looking at a new dataset. Dr. Peng does a great job walking through the steps and there is enough information given to enable the student to effectively explore on their own.

The course was really good, thanks for that; however the part of single value decomposition and principal components analysis was not explained in a gradual fashion and even though I researched outside of the course I still have some confused concepts there.

Good, but the lack of assignment in week 3 seemed to screw up the UI, prompting me continually to do the Swirl exercises, which were non-compulsory (and, given I hadn't completed any of the other Swirl exercises, something I didn't want to take on.)

The only missing point I would say about this course is how to deal with skew data and/or outliers. Although it is not specific to "cleaning data", I think there is a good opportunity there to at least give some hints on this subject

The Swirl activities followed along with the lectures, which allowed us (as learners) to better understand core concepts. The lecture videos continue to end while the professor is still speaking, but this is not a major issue.

It was awesome to learn visualization. SVD and PCA part of the course could have been elaborated better, and a pilot project on that would have cleared the basic concept. As usual Prof. Roger is a engaging and amazing teacher.

The course was great, I'm not sure if I'd really consider using the base plotting package in reality as the plots are just too ugly, and the API is harder to learn. I think a stronger on ggplot would help to keep it relevant.

I enjoyed the course and learned important graphing concepts for R/RStudio. I just wish the assessments had been a little more rigorous, as it felt like I could have done better but still passed the projects anyway.

This is a very good course, at times it felt like the instruction was to do things mechanically without understanding the motivation. Perhaps this should come after or in conjunction with Statistical Inference