- 4 reviews
- 4 completed
The course is structured such that you're given a coding challenge and after you attempt a solution Peter will give you his solution. While your code may work invariable Peter's solution is far more elegant and succinct. This is the best way to learn: try something yourself and only after you've struggled with it do you get alternative (and usually) better solutions. This is pure beautiful coding. You're building for simplicity and ease of understanding. Nothing overly complicated. If you like militantly following rules and being told exactly what to do this course is not for you. If you're willing to explore on your own and learn from a master coder it is.
I took this course in the Fall of 2011 and it's one of two courses that inspired me to create this site (CourseTalk). Here's the best things about this course: \- Andrew Ng is awesome. He's a top expert in the field and you really feel like he's your personal tutor. \- The course makes machine learning very easy to understand. \- High production value.
Taking this course from Sebastian was a pleasure. To start with he is exceedingly qualified to teach it. He was the director of Stanford's artificial intelligence lab and his team was the first to complete the 2005 DARPA grand challenge with their self driving car Stanley. He went on to lead the development of the Google self driving car. In this course you'll learn the core algorithms that power Google's self driving car. And you'll learn directly from the source. "That's pretty cool!" as Sebastien would say. The passion and enthusiasm that Sebastien has for the subject comes across in the videos and is infectious. The course is broken down into short video segments generally not more than 5 minutes. Plenty of quizzes and programming assignments are dispersed to keep you engaged and make sure you're learning. Growing up I was a huge fan of Legos and my favorite thing happened to be building cars. I love machine learning, I love designing things, I love building things, and if you're like me you'll love this course. My idea to make this course even cooler - create some kind of virtual world where students can deploy their own code to test in a simulated self driving environment. Perhaps even create a race or competition. Or alternatively, provide instructions for creating your own miniature self driving car for testing in your living room.
I came at this course without any formal background in computer science. Everything I know I've learned as a result of simply learning as I go. However it occurred to me I could never pass a technical interview because I didn't know the most basic of algorithms that interviewers expect you to know. Pros: \- Taking this course made me feel like a bad-ass software developer. I now have a good understanding of algorithm complexity and how to develop solutions in any areas where performance is an issue. \- Tim knows his stuff Cons: \- Tim's handwriting is often quite poor AND what he is diagramming/writing often seem to be way out of sync with what he is saying. This makes things really confusing. I often found I had to look away from the screen and just listen to his voice. \- In some of the other courses I've taken I really get the feeling the professor is there to help me learn.. In this one it feels like Tim is there to simply convey information. \- This course did not take advantage of the medium. For algorithms like quick sort or merge sort I can fully understand it with a simple animation so much easier than a long description. See here: http://www.youtube.com/watch?v=y_G9BkAm6B8 In conclusion I think with a bit more effort this course could be amazing. Having said that I did learn a lot and found this to be a very valuable course.