Avis et commentaires pour l'étudiant pour Exploration analytique de données par Université Johns-Hopkins

First two weeks are too repetitive with other courses

I would have liked an assignment to focus on the clustering methods and I think dimension reduction was reviewed way too quick.

The course content applies to R. The teachers focused on the programming language rather than the application of the existing graphs to explore data.

I wish there several comprehensive examples of exploring some real data as guided by the course instructors.

The course covers very limited subset of plots and mostly oriented to R-specific technical routines rather than overall approaches. Case-study example is helpful and contrary to the most comments I do appreciate the final course project: this how most problems are stated in real life. If you would like to cover more fundamental concepts behind exploratory analysis I would recommend other sources.

A lot of broken swirl(), which wouldn't be so bad except *a lot* of this course is based entirely on swirl(). Also the swirl() text was almost verbatim of the lectures one has just watched.

All in all, good information, but the swirl() badly needs an update.

Course feels somewhat abandoned by instructors. Content is okay, but can't help the feeling that it's basically a cash cow - students would benefit a lot if instructors were move involved.

Good course, but the SWIRL exercises (and a few quiz questions) needed to be updated for the latest version of ggplot2.

Good course overall. There tends to be many lectures that are just lists of commands. Also, they don't seem to be updating the material. Many lectures are several years old and still have typos in them.

Good class - links and slides have not been updated recently. Frustrating to finish without the exact linkts to the data.

The course was overall ok, but I wish discussions about k-means, PCA and SVD were divided into two courses.

did not like the swirl-tutorials. they were very tedious and sometimes labelled correct commands as false (e.g. when you typed = instead of <- for assigning a value to a variable)

also I was surprised that for a beginner programming course in R you had to apply specific functions such as grepl without the function being introduced in the course

Very unprofessional, compared to other courses. It wasn't well organized.

Veyr boring

Some parts of material is good quality, but some is bad - also some show bad practices in R. Extensively use swirl as assignments over self work. It is better to go through good tutorial over R base plotting system and ggplot2.

too superficial

About plotting in R. Not about generating real insights from EDA.

didnt get much useful-- a whole "course" on plotting? meh.

Clustering topic is covered superficially, too much time spend on employing ggplot graphs, not very useful since making graphs is straightforward on other software, like excel, once you aggregate datasets correctly. I had not found it very enriching as a course. I would merge this class within R-Programming section and call it Part 2 rather than categorizing into "Exploratory Data Analysis".

Alas, after only 10 minutes of the first video, I am reminded that this instructor does not gear his lectures to the true Beginners among us. He speaks much more for an audience of grad students. I do want to complete this Specialization, so I will try again perhaps after learning more - about statistics and R and who knows what else. I fought my way through the first three courses, but now I'm going to work smarter by finding other ways to acquire this knowledge. Then return to him; maybe. This course should be labelled Intermediate and Statistics should be listed as a prerequisite. (I think; since I don't know what it is that I don't know, I am making a guess as to the missing piece of the puzzle.)

Course content and assignments were difficult to follow. Loads of statistical content along with high-level R content means it was probably the toughest of the 4 I have taken so far in the Coursetrack.

some swirl tests (4,5) don't work because of parameter method in qplot function. This parameter is not realy existed in this function now

The final project did not require use of the material in the course beyond the first week and a half. I did not take any quizzes or otherwise have my knowledge tested on the material in the second half of the course.

Quite repetitious in covering basic graphing, and very shallow in regards of clustering, SVD and PCA.

Absolutely No technical help, like insane amounts of homework for each week. People have jobs and businesses to run. Incredibly short duration. Like literally this should have been spread out several more weeks. I would have dropped the class but I can’t. It’s so difficult to get i to the first set of practice assignments and these several sets. Honestly, I am literally getting no help on it and probably won’t pass because I am missing the deadline. I finished 5 coursera courses working on them for 24 none stop. I’ve literally been at this class all day. Besides all the insane amounts of assignments there’s tons of videos to watch and uploads to do. Go buy some books or take another class unless you are unemployed or have nothing better to do.