AI Programming with Python

Course needs to be improved. Unsatisfying. Too much too fast.
I am about 90-95% through this course. I have some experience in coding, but am very rusty. I have a Masters Degree in Mechanical Engineering, so I do have a technical background

I took this course intending to learn Python before getting into AI/Machine Learning.. Would probably have been better to take specifically a Python course.

There was a lot of Python in the beginning, and it seemed like it got very deep in a hurry and expected the learner to be pretty proficient. Then, poof, no more learning Python. The course was really 80% learning Jupiter Notebooks, Pandas, Seaborn, Numpy, Matplotlib, Pytorch, etc etc etc.. There were several videos on calculus and linear algebra (for example) that I really think could have been cut down 90%.

There was some discussion and exercises on backpropagation, and the course emphasized how important this was... but, i really felt like they could have done a better job walking the learner through this. I don't feel that confident with the Python implementation of this important concept.

I was able to understand conceptually what was being taught through the videos and the notes. But it was difficult to apply what was presented in a lot of the exercises, quizzes, labs and projects.

The course seems kind of thrown together. Some of the instructors are very repetitive. And they overlap. They showed 4 or 5 times a cartoon guy climbing down a mountain with a funny name using that as an analogy for gradient descent. OK, WE GET IT. But, then, they quickly gloss over the specific variables in the coding exercises, and the learner becomes lost in what is going on. Gradient descent is suddenly called an "optimizer" ... huh?? When did that happen? There are several examples where the learner is left trying to figure out what the course means when they start calling things different names in the code.

When they aren't being repetitive, they tend to blow quickly through things that catch you in the exercises. They really do a poor job, in my opinion, of really explaining clearly everything that is happening when they throw you into these exercises.

The first exercise was very confusing- the "TODO"s really weren't clear.

I made use of the "online mentors". Turnaround was not under 24 hours. By the time the guy came back to me, I already had wasted too much time on the issue I was having and moved on. For me, not being a CS major/software person, I found it exceedingly hard to get help in this manner.

This is kind of small potatoes, but to me, it is indicative of poor editing and hasty work- there are a lot of typos and grammatical errors throughout. In some places the course points this out, that there are errors in the videos.

The videos in a lot of cases are not really that helpful in my opinion. They go much too fast and cover too much. I watched every single one anyways. But really, it was the course notes that followed the videos that was most useful.

I am close to the final project, and unfortunately, I am kind of dreading this.

On top of this, I always felt the clock ticking in the background... I felt I had to rush through this material. Very challenging to do this in 10 hours a week. (Work, family, etc.)

Overall, too much, too fast, too disorganized and inconsistent. Not really that coherent. This course needs to be redesigned.

I have gotten some familiarity with neural networks through this course... But, I wouldn't be able to put together a neural network on my own today.

Also- there ought to be a pdf available for the customer to use as a reference. For $400 you get, like, 6 months access to the material online.

It started out great - Having little experience with programming or python, I really appreciated the first lessons and videos. I made progress. I invested a lot of time. And still the recommended timeline seemed extreme. I guess they realized that as they had to postpone the final submission deadline two times.

But then, as many have pointed out before - it got very steep, way too fast. Nothing was explained, you had to somehow figure out what things meant - i did not understand how this jump from "hey kids - this is a variable! Let's look at some examples together" to "now do your own project - you'll figure everything out on your own! You've got 24h hrs" happened.

There are alot of mistakes within the lessons but somehow they never cared enough to correct those. This course has been around some time... So this seems lazy and careless to me.

I would not recommend it to anyone without some kind of in debth experience in AI programming. Those of you can skip the first half of the course and jump to the projects. But then I wonder why you would need that course in the first place.

I am kind of disappointed as I have spend countless hours, and nights on this course to - at the end - realize that I am nowhere near to being prepared enough to finish the last project. I spend months!!! working on the last project - None of the help given got me further, I got so confused and lost in this vast topic with nowhere to hold on to, that I am now in peace with giving up. I am not going to spend any more time, working on something that has stopped teaching me anything months ago and since then just confused and demotivated me.

Overall, I am very pleased with the program's progress. I anticipated that I would need to dedicate extra time to learn math subjects crucial to success in AI and programming, and the courses have effectively broken down complex topics into manageable segments. The session lead, Raju, has been an exceptional teacher and mentor. He ensures that all students are making progress and has accelerated my learning by providing guidance on areas where I need to focus. The program has exceeded my expectations by providing hands-on projects that build my skills and confidence, particularly in learning Python and math. The course's realistic expectations have motivated me to continue programming outside of the course. I highly recommend this course to anyone willing to put in the effort to learn, as the supportive and talented community of both students and teachers are incredibly motivating.

All over the place
I'm actually a big fan of Udacity, and I've done two other nanodegrees. However, I felt like this course was all over the place, they would do some basic videos and instruction on simple, fundamental things on Python (like 'this is a class', 'this is a function'), and then the next practical exercise would be a a 60-step process involving tons of libraries and intricate steps that would be nearly impossible to follow along with unless you're very familiar with Python. In general, I felt like they put together a bunch of topics in AI that don't really blend together that well, and tried to make a class out of it. I was half-way through the last project but spending so much time on it without learning that much, so moved on to other things. It might be worth coming back to at some point, but kind of steep price for a course that is not as effective as datacamp or Udemy.