Learner Reviews & Feedback for Data Science Capstone by Johns Hopkins University

I enjoyed the course. This course took me waaaay more time than I thought because I struggled with a few issues. First, I wish I'd started by taking the NLP online course before starting the Capstone (https://www.youtube.com/watch?v=-aMYz1tMfPg). There was an issue installing RWeka, RJava and it took me several days to work through the issues. I eventually moved to using quanteda (https://cran.r-project.org/web/packages/quanteda/vignettes/quickstart.html). I also waited far too long to develop a method to test my model using a subset of the training data, so I could test whether changes to my model improved and reduced performance. It turns out that my model trained on a 25% sample performed just as well as a model trained on 100%. I'm thankful for the Discussion Forum and final peer review process. Both helped me learn how I can improve my model and demo application. I really appreciate the instructors for creating this specialization. I've learned a lot.

A very tough and challenging project, but a great way to learn a lot about Natural Language Processing and algorithm coding in R, and in the end to have a cool Shiny app to add to your portfolio. The project weekly structure could be enhanced (maybe adding one more week could help) and the weekly instructions, while informative, could also be improved. Thankfully the forum has been very helpful. Informative and motivating videos but where were the SwiftKey people mentioned ? Finally, the quizzes 2 and 3 should be replaced by other exercises with more educational value. Overall an interesting learning opportunity !

Interesting project. Not too tough if you take it slow and simple: my first version was already quite good and I earned a full score without going through the weird circumvolutions I saw other students get into.

New subject I loved to learn about.

It is a bit disappointing though that this does not follow the complete 10 classes: in my case, the knowledge I learnt in "Statistical inference", "Regression Models" and "Machine Learning" was not applied in this project, whereas we could say those courses are the core of the specialization.

This is not for beginners with no experience. The estimated weekly hours are absurdly low.

No one has seen any sign whatsoever of the industry partner, SwiftKey, despite claims they will be around to help. The field has advanced dramatically since the course was developed. Be prepared to do a lot of research and trial and error.

The specialization has been an excellent way to learn a fair amount on the topic, but it is just the beginning. The capstone will challenge you. It is rewarding when you complete it.

With this course I learned to go through a data problem from the scratch and get a real data product, and document it. My only constructive comment is that, when reviewing the projects of classmates, there is a huge dispersion on the effort and quality of the products presented, from very basic and simple Apps to a very professional products, and the scoring of them all is quite the same, perhaps one or two points of difference, in eleven points maximum score.

Feeling proud after completing all the courses under Data Science Specialization. This was not an easy task to complete especially if you are not familiar with the Statistics. Requires continuous dedication and motivation to follow and complete. Course is well designed and cover most of the topics. Its just stats part can be enhanced further to cover some basic aspects. Thanks for all the support

The course was extremely challenging. It was an excellent capstone to close this full specialization. However, just a little bit of more guidance regarding language models and n-gram models would have been perfect to make the journey easy. However, there's a lot of value on researching on our own and finding the right answers without spoon-feeding. Thank you for everything!

its good learning. however NLP is never introduced in the course. its very hard to compute the metrics and train the model in R. Accuracy of the model is very low, as it's not easy to train the models in R as it requires humungous data sets to get up to 95% accuracy.

Neural networks, LSTM in python with tensor flow is the best choice to do the projects like this.

Good scenario and a good learning opportunity. I don't think the quizzes related well to the problem we were trying to solve and introduced a red herring, however. Predicting the next best word is not the same as predicting the relative probability of 4 words where one is the "right answer" but not necessarily the best prediction of a text prediction algorithm.

This is a great capstone project. It requires the student to really have an understanding of the concepts learned throughout the specialization and apply them to build a prediction app. There is very little guidance asides from the discussion forum which could be discouraging at first. Otherwise I would've rated it 5 stars.

Overall a good course that makes you learn a lot on your own (unlike the rest of the series). Maybe a bit too much self learning. However, if you can complete it does give you a lot of learning especially in some text analysis which hasn't been covered before.

The course project was very interesting. It can be challenging if you want to do it properly or easy if you just want to pass. I tried to do it properly for which I had to repeat the course 3 times, but in the end it was good - I think I learned a lot.

The Capstone Project makes you summarizes what you have learnt so far and take it to the next level, natural language processing . Besides, the ability to create a working app is a reward by itself.

It is challenging and interesting at the same time.

The capstone will be much easier if participants have hands-on experience and understanding of how R works. For those who have managed to overcome the steep learning curve, the reward from the learning experience is well worth it.

It was a challenging project and really pushes you to learn and manage on your own. It also pushes you to build and end to end product within time and memory constraints. Learned a lot during this project!!

Thanks!!

its a bit challenging "introductory" course to Data Science. I like it but some weeks has too much info before a quiz or practice. in general is a good course, better than google data analytics

It kinda feels like you versus the world in the first stages, but after correctly implementing and answering the questions poised by the videos and lectures it all stats to click, keep focused!

The content of this course is very good and the assignments test the knowledge gained

The video lectures are some times boring, losing focus.

I think video lectures need some improvement

It is a lot of independent work, with guiding questions but no real help otherwise. If that's not your thing, this course is not for. If it is, you'll great a very fun end project.

I appreciate all the work they put into creating the course,. However, it can be frustrating to follow. It would be nice if they would structure it in a more organized fashion.

I wish that either there were a choice of capstone projects, or that there were a more numerical component to the analysis than such a pure text based assignment.

Very good Course as a beginner course for Data science , you will learn a lot of stuff and the capstone is a very good starter for Natural Language processing

Very interesting, but Capstone is much more difficult than the rest of the course.

Be sure to study carefully the first courses, and don't rush.

I wish we had more training on text mining. Was not that easy! But overall the program is very interesting and challenging! Enjoyed it a lot!

The project is really interesting by itself, but there is a lack of preparation and instructions to build it, basically you are on your own.