Algorithms on Khan Academy

Khan Academy recently announced they published an introductory module on Algorithms. While I am waiting for the "Principle of Computing" and "Algorithmic Thinking" classes on Coursera to come around again, I thought this would make a nice preparation step. I am just under halfway through with the module.

What is great:
This is a very nice introduction to algorithms. While I don't feel like I could walk into a job interview and code these from scratch, I do feel like for every one of the algos covered so far, I can give a solid description and pseudo-code to describe how they work in general.
I also liked the structure of the algorithm sections: introduce the algorithm, give a part of the implementation, and then the rest of the implementation, and finally analyze the efficiency of the algorithm.

Not so great:
The implementation sections have an specific, expected solution and gives "hints" to this solution. You won't get credit for a solution - even if all of your tests pass - until you match exactly what they "hint" at. Personally, I would find it more effective if they accepted a working solution then guided us to a more elegant solution. Moreover, there are instances in which my code matches the hint, but I am still not getting credit for it. It is just unclear if I need to do more or if there is a bug. In these cases, I wish the algorithms section was more like the rest of KA, in that there are multiple hints that increase in their help.
I also thought the section on asymptotic notation could have been clearer. I would get the right answers on the quizzes, but I wasn't always sure why that was the right answer.

Overall, this is a really good introduction. I think after working through this, a novice like me will be able to better tackle some of the more advanced algorithms courses on Udacity or Coursera and have a nice little head start.

Principles of Computing - Abandoned

I dropped the PoC class. I just came to the realization that I really needed to bone up on Discrete Mathematics before going any further in Computer Science. I found myself increasingly out of time every week, and cutting the math-heavy homework to focus on the mini-project. I'd rather be able to do both and really deeply understand the material than just trying to scrape a decent grade.

I really liked the idea of using the MIT courses, but without a set of answers to the problem sets, it is not clear to me how useful it is. It seems like such a good course too. The recitations and the exams *do* have solutions posted. Perhaps I can find exercises to fill in the gaps.

This seems to be my only option, as Coursera doesn't appear to have anything that will fill my needs any time soon. Doing a quick search isn't turning up any alternatives of the same caliber of the MIT course. Bleh. What is frustrating is that there does appear to be a site with links to solutions, but all password-protected. Rats.

This site had some suggestions ofr courses as well as books. I thought this Discrete Mathematics Book by Epp looked promising, but for $200, I'm going to have to think seriously. Amazon does allow for rentals. Maybe there is a cheap version from the library? The other books in pdf on this site needed Haskell. Not sure I want a programming and mathematics combination.

Hm. It was a lot easier at university. Classes were pre-determined, and all anyone had to do was sign up, pay the crap load of money for the class and books and fees, and then make it through. Piece of cake :)