danger

9 ms·

31.
by danger 16y ago
One other issue that comes up is "garbage time". When a game isn't close (say in the last quarter of a blow-out), the stats are basically meaningless. Does Pomeroy have good ideas about how to deal with that?
32.
by danger 16y ago
I haven't run it yet. It should be done in time to enter the contest, though (but it will just be a baseline--i.e., not eligible to win prizes).
33.
Selection Sunday: Is your March Madness prediction algorithm ready?(blog.smellthedata.com)
10 points by danger 16y ago|13 comments
34.
by danger 16y ago
Thanks. I (obviously) agree. The only problem is that it's not always computationally easy to optimize the thing you really care about.
35.
Predicting March Madness: On Tournament Structure and Bracket Scoring Rules(blog.smellthedata.com)
8 points by danger 16y ago|3 comments
36.
Artificial intelligence scientist gets $1M prize(cbc.ca)
2 points by danger 16y ago|0 comments
37.
by danger 16y ago
This paper (referenced in the post) is also relevant. The application is to NBA basketball: "Incorporating Side Information into Probabilistic Matrix Factorization Using Gaussian Processes." Ryan Prescott Adams, George E. Dahl, and Iain Mu
38.
George Dahl: Machine Learning for March Madness(blog.smellthedata.com)
25 points by danger 16y ago|3 comments
39.
by danger 16y ago
There are many different actual tasks that technically are PASCAL challenges, but when people say "PASCAL VOC challenge" (Visual Object Classes), they typically mean either the _classification_ or _detection_ challenge: Classification: For
40.
by danger 16y ago
Another question: are there other scenarios outside of playing Scrabble where something like this would be useful?
41.
Machine Learning for Human Memorization(blog.smellthedata.com)
40 points by danger 16y ago|4 comments
42.
An Algorithm to Generate Impossible Art?(blog.smellthedata.com)
3 points by danger 16y ago|0 comments
43.
by danger 16y ago
Update: finally a good answer. If you want to do the people at stackoverflow a favor, upvote this answer: http://stackoverflow.com/questions/3956478/understanding-ran...
44.
by danger 16y ago
If we are interested in a _sequence_ of pseudorandom numbers, then we should be talking about the _joint_ entropy of the sequence. But information theory is plenty capable of describing "how random" a sequence is. The generator you are tal
45.
by danger 16y ago
Or you could argue formally by pointing out that the maximum entropy distribution on the interval [0,1] is the uniform distribution, so modifying that distribution will only decrease its entropy, which means that it's "more predictable". h
46.
by danger 16y ago
How is that related to the SO question? The post is asking about how to tell if one distribution is "more random" than another, which is what entropy is all about.
47.
by danger 16y ago
In what sense is a "fuzzy graph" different than a standard weighted graph? http://en.wikipedia.org/wiki/Weighted_graph#Weighted_graphs_...
48.
by danger 16y ago
Among others, I liked this quote: This is not to say that counterfactual thinking is not encountered at all outside of mathematics. For instance, an obvious source of counterfactual thinking occurs in literary fiction, particularly in spec
49.
Counterfactual Thinking(terrytao.wordpress.com)
5 points by danger 16y ago|1 comments
50.
by danger 16y ago
Why has nobody in this thread or the stack overflow thread linked to the Wikipedia page on entropy? http://en.wikipedia.org/wiki/Entropy_(information_theory)
51.
Fundamental Limits of Locality(blog.smellthedata.com)
6 points by danger 16y ago|0 comments
52.
by danger 16y ago
To add some context, this is the paper with all the details that wasn't available when the earlier press release hit HN. The old conversation is here: http://news.ycombinator.com/item?id=1732952
53.
by danger 16y ago
Glad you like the links, but I disagree with your disagreement =P You can use factor graphs for either directed or undirected models. See this paper: Extending factor graphs so as to unify directed and undirected graphical models B. J. Fre
54.
by danger 16y ago
I use them, but there are plenty of libraries out there. http://compbio.cs.huji.ac.il/FastInf/fastInf/FastInf_Homepag... http://robotics.stanford.edu/~sgould/svl/ http://code.google.com/p/factorie/ http://people.kyb.tuebingen.mpg.de/
55.
by danger 16y ago
I don't think that's true. See, for example, Finding Maximum Flows in Undirected Graphs Seems Easier than Bipartite Matching (Karger-Levine, 1997): http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.52.3...
56.
by danger 16y ago
The method sounds very interesting, but the title seems to overstate the work a bit. This is an approximation algorithm, meaning that it's an improvement on an _approximate_variant_of_ a fundamental problem, right?
57.
by danger 16y ago
There are so many possible directions that somebody might want to go, though. If you're ultimately interested in reinforcement learning, you don't immediately need to understand EM in order to "get" Q learning. Or to work on discrete prob
58.
Choosing a First Machine Learning Project: Start by Reading or by Doing?(blog.smellthedata.com)
30 points by danger 16y ago|6 comments
59.
Help Create a Machine Learning Stack Exchange(area51.stackexchange.com)
5 points by danger 16y ago|0 comments
60.
Bayesian NBA Basketball Predictions(blog.smellthedata.com)
6 points by danger 16y ago|0 comments

More ›