From time to time I like to find an existing project on the web and try my best to either improve the app or to make a better version of my own. I’ve even hosted contests in the past for developers to do the same. During those competitions, I’m almost always asked where do I find applications that are worth checking out.
Related Posts
With the Heritage Health Prize beginning last night, the majority of my time is going to be spent on that for a while. If any of you are participating in the $3 million prize contest, I want to wish you all the best of luck. I know when I began working in the machine learning industry, one of the biggest problems I faced was how and where to obtain a good collection of data to design and test with. For those of you that are not (and those that are) participating in the contest and would like to do some work with algorithms, machine learning, and neural networks and are looking for some good datasets for developing and testing with, I want to share a few of my favorites with you now.
Related Posts
Alrighty. Today is the day that the Heritage Health Prize begins. Good luck to everyone. And, wish me luck as well!
Related Posts
In September of 2009, Netflix awarded a grand prize of $1,000,000 to the “Bellkor’s Pragmatic Chaos” team for their creation of an algorithm that predicted user ratings for films based on previous ratings. The new algorithm was an improvement over Netflix’s own algorithm by 10%. The reason Netflix awarded a prize of $1 million was because they felt that it would be easier, faster, and cheaper to hold a contest that allowed some of the world’s best mathematicians and programmers a chance to develop this much needed algorithm instead of attempting to develop the algorithm on their own. The contest was a huge success. Unfortunately for me, I wasn’t made aware of the contest until after it was over. Yeah, I know. I was living under a rock. 
However, the good news is that other companies are spinning their own contests in the same nature as the Netflix contest.
The Heritage Provider Network (HPN) has setup a contest of their own. The “Heritage Health Prize“, as they’re calling it, is a contest to design an algorithm that uses available patient data, including health records & claims data, to predict and prevent unnecessary hospitalizations. Unlike the Netflix prize, the Heritage Health Prize carries a hefty $3,000,000 purse. HPN states that more than $30 billion is spent every year in unnecessary hospital admissions. They also believe that an algorithm, such as the one in their contest, could prevent these unnecessary hospitalizations, saving the majority (or hopefully all) of that annual $30 billion.
Lucky for me, I am aware of this contest and will be participating in it. The contest begins tomorrow (April 4, 2011) and is expected to run for exactly 2 years. If you are interested in competing in this contest, you need to head over to http://www.heritagehealthprize.com/ and register. Today, there is only a “pre-register” form that is located in the upper-right corner of the screen and only asks for your email. Should you decide to pre-register, tomorrow, you will be contacted with the official rules of the contest along with more information included how and where to obtain the datasets for designing and testing your algorithms. Since I will be participating in the contest, I’ll come back here and post as many details and updates as I can without giving away too much information or too many details about our algorithm itself.
For any of you that decide to take a shot at this contest, I want to wish you all the best of luck! Plus, I would also like to ask each of you to come back here and let us know how things are going for you, your team, and your project. Feel free to post questions in the comments below if you would like for assistance from other users. Since I am personally taking part in the contest, I won’t be able to help you as much as I could normally. But, I’d still be glad to offer assistance where possible and I know that our readers can also offer a ton of expertise if you ask. So, again, good luck to all of you and I hope to see one of you be $3 million richer in 2 years!
Related Posts
A few weeks ago, I bought several books from Amazon.com to add to my library. Of those books, I’ve read 3 of them and I now want to take a minute to review the first of those 3. The book I want to talk about today is called “Algorithms of the Intelligent Web“. The book was written by Haralambos Marmanis & Dmitry Babenko and was published by Manning. It is a hands on book for learning some of the biggest algorithms used on the web today (hence the name). The code examples are written in Java, but are not complete in most cases.
I am a huge fan of anything related to artificial intelligence and any kind of algorithm. So, I was excited when I came home and found this book in a box with the other algorithm books I ordered at the time. Unfortunately though, I wasn’t as pleased when reading it. I like that the book uses real-world applications for its examples and that it uses Java as its coding language. But, this book was not up to par with what I consider a “good read”. The book was extremely difficult to follow and seemed like it cut you off with no where to go.
Every section started off the same way. The sections would begin by showing you the code that’s going to be used throughout that section. However, by the time you get to the end of the section, you find yourself having to flip back to the beginning of the section to re-view the code. This became extremely annoying right away. Plus, as I’ve already mentioned, by the time you get to the end of the sections, it seemed like the sections would end abruptly and never really finished anything the way it should have. At the end of each section, I was left with more questions than answers. I’ve read several books just like this one, and all of those did a better job at answering questions. For example, the book “Programming Collective Intelligence” discussed the same exact algorithms that Algorithms of the Intelligent Web did, but Programming Collective Intelligence did a much better job of explaining things and was an overall much better read.
Algorithms of the Intelligent Web wasn’t a terrible book, but I would have to say that there are better alternatives. As in the case of Programming Collective Intelligence, all of its examples were written in Python. But, the text itself did such a better job at explaining the algorithms that even if you’re not a Python programmer, you can still follow along without any problems. However, if you absolutely have to have Algorithms of the Intelligent Web, you can grab a copy at Amazon for a pretty good price if for nothing else but to have it in your library.
