Didn't Feel Like I am learning some concept very basic concepts nothing related to real life and NLP

In general, good. M​isspellings in assignments is not very professional, should be revised.

Would be good if there are more checkpoints to see if the codes are correct or not.

Some example is hard to understand, and few of the diagram is ambiguous.

Especially later excercises contain code/instructions that are unclear

I would like to have more content and explanation about the math

I don't feel the assignments help understand the material.

The videos are too short. Discussions are oversimplified.

It is an interesting course to learn the basics of NLP

good for beginners, but needs more advanced concepts.

The explanations in the videos could be improved.

No longer required. Beyond my present knowledge.

Content is verbose and locks context in places

I believe that explanations where too shallow

v​ery good course, a lot of stat and math

So many small mistakes here and there

I didn't quite get the feel of it

Problems in week 3 Assignment

Just ignore the video!

Too easy

Too easy

I believe the course does not allow you to study NLP in depth. Compared to the deep learning specialisation by deeplearning.ai, this course has probably hours(!) of video material less. PCA is for instance presented in ~4 minutes and the lecturer concludes with "now that you know all about PCA". The only further reference provided is a link to the standard textbook in the field, no detailed study guide or references for individual topics. Excercises are done in notebooks and test beginner python skills instead of nlp understanding (Basically: "Look up key i in dictionary j and store vector k"). It does do a good job in giving an overview about NLP.

I was torn between two and three stars. Two, as mathematics are dealt with far to shallow. No proofs, no motivation, nothing. And in the final week, there is a massive notebooks with a lot of flaws and a lot of cells you have to code in a specific, sometime suboptimal, way. Otherwise, the grader will throw errors. All in all, things are handled to shallow and it is more of a coding lesson than a deep dive into ML, which necessarily requires mathematics. This is one more of those "Become a data scientist without mathematics" things the world does not need.

lectures are pretty mediocure. basically it lacks motivation behind algorithms, you are simply told what to do, really like "machine" learning

you'll spend a lot of time in the assignment, not focusing on implementing your algorithm, but adjusting incorrect input or output format that passes all tests but fail the final grading for some reason (also in week 3 the assignment has one or two questions that do not even tell you what's the input data and you have to "print" them to get an idea lol)

The lectures skip over important features that should be explained in more detail. Other important concepts are left to the labs, even if those require a good explanation. Evaluations are not a help to practice or understand concepts. Most of the time spent on evaluations is figuring out how to do things in Python rather than follow the concepts. People can obtain 100% in the evaluations but learn nothing. It is indeed a very poor course.