|
I have tried to implement this programmaticly since I first read about it in the late 1990's with zero success. The method for learning this is Phenomenal (with a capital P). Now I have been going back to many failed attempts and seeing old code finally run! Exciting. My question is am I alone? Does anyone else thing that this method of instruction has helped them while trying to learn the same subject in the past did not work. I have a reason for the question, but lets just say it is curiosity for the moment. |
|
Most definitely. I didn't understand particle filters until AI class + CS373. I was particularly confused about how to go about sampling from a distribution, but now I do! Many other concepts were further clarified as well. |
|
The concepts are new to me so far in this class. But I had that same experience in the ML class (Coursera); the concept of back-propagation in neural nets finally became very clear to me. It feels wonderful when old confusions are cleared up like that. These classes are a lot of fun!! |
|
I too remember trying to learn A* before with little success. It was back in high school (how long ago was that again... about twelve-thirteen years ago, I think) when I wanted to try and make a strategy game... Back then, I never quite got it, and ended up making some different games that didn't require pathfinding :) Now, with the help of Udacity, I finally got it. And it wasn't even that hard. I think it's a combination of Prof Thrun's great teaching skills and the method used on these courses: short snippets of video followed by a quiz. When you can't move on without understanding the concept, you have no choice but to lean ;) |
|
The AI class was fun, and I thought I understood the concepts, but it was not until this class, where we actually implement everything, that I really understand the particle filters, search algorithms, etc! |
