jmount

77 ms·

871.
Estimating generalization error with the PRESS statistic(win-vector.com)
2 points by jmount 12y ago|0 comments
872.
by jmount 12y ago
That is something to be careful about with our book. We largely use machine learning- not tinker with or develop it.
873.
by jmount 12y ago
One of the blog/book authors here (John Mount). Yes the book and blog are both by Nina Zumel and John Mount. Obviously my opinion is biased, but here is a bit anyway. "R in Action" is one of our favorite R books (that and
874.
Factors are not first-class citizens in R(win-vector.com)
27 points by jmount 12y ago|17 comments
875.
by jmount 12y ago
And the prof married one of the involved graduate students. (no: I don't think marrying a graduate students is in general wrong- but marrying one you supervised in an abusive experiment is a bit off).
876.
by jmount 12y ago
Ph.D. programs vary a lot by country, school, and department. They mostly appear uniform from the outside. Big variabilities include: funding, status, teaching load, qualifying exams (these are after you are in), amount of choice, publica
877.
by jmount 12y ago
That is why you have to check. If the P(data|param) is concentrated, then even for uniform priors on the real like you can have P(param|data) is proportional to a sensible distribution. But you have to check for specific model and data (s
878.
by jmount 12y ago
I'd say the justification is two parts. First is the Bernstein–von Mises theorem (priors don't matter once you have enough data, as long as you didn't violate Cromwell’s rule by using zeros). The second part is: improper pr
879.
by jmount 12y ago
Nice article. In this direction my group has been trying to help teach that you tend to need to be familiar with both frequentist and Bayesian thought (you can't always choose one or the other: http://www.win-vector.com&#x2
880.
by jmount 12y ago
Site has a completely unusable user interface.
881.
Half of a Coin: Negative Probabilities [pdf](wilmott.com)
60 points by jmount 12y ago|38 comments
882.
Fractional calculus(en.wikipedia.org)
9 points by jmount 12y ago|1 comments
883.
Women in Data Science – Positive Imagery for the Win(ninazumel.com)
4 points by jmount 12y ago|0 comments
884.
Automating variable treatment for modeling(win-vector.com)
1 points by jmount 12y ago|0 comments
885.
Vtreat: a simple library for preparing R dataframes for analysis(win-vector.com)
1 points by jmount 12y ago|0 comments
886.
by jmount 12y ago
The backprop idea is one instance of a general idea called reverse accumulation (which you can use in other contexts): http://www.win-vector.com/blog/2010/07/gradients-via-reverse...
887.
by jmount 12y ago
Not your comments, just the occasional comments snarks of others (sorry, I don't actually have a specific example at hand). Your comment is actually helpful (mentions a change has been made, explains the decision, and even mentions th
888.
by jmount 12y ago
The original researcher spoke out and complained, so I it is safe to assume they to some degree consider discussing this helpful.
889.
by jmount 12y ago
The thing is: there is no Solomon/neutral solution to this. Either the original researcher was harmed or not. I feel a need to be fair to the original researcher as well. This entails calling out people who harmed them, or allowed
890.
by jmount 12y ago
Anticipating the web-site not being able to take the load is why I linked to Boing Boing (also the slightly less provocative title). The thing I mind about titles and links being changed (which probably is in fact a good idea) is the commo
891.
by jmount 12y ago
This issue isn't expecting new research from science fairs (which would be a bit over the top), but claiming something your parent published in 2011 is your work and novel in 2014. But yeah, at least it wasn't a papier-mâché volc
892.
Sixth grader's internet-famous science project may have been plagiarized(cflas.org)
71 points by jmount 12y ago|39 comments
893.
by jmount 12y ago
Rays usually means em-radiation. But cosmic rays are misnamed from (as they tend to be radiation, but not rays).
894.
by jmount 12y ago
But not typically at the so-called "cosmic" energy scale.
895.
by jmount 12y ago
I thought solar flares interacted in ways that suppressed cosmic rays (such as http://science.nasa.gov/science-news/science-at-nasa/2005/07... ), not produced them. But it isn't my field, I could easily
896.
by jmount 12y ago
One flaw that bugs me in the article: "the dangerous cosmic radiation spewing from the sun in the form of solar winds." From the Wikipedia: "Cosmic rays are immensely high-energy radiation, mainly originating outside the Sol
897.
by jmount 12y ago
Wow, that is really ugly semantics. Here are some notes of mine on how hard R works to avoid exposing this sort of aliasing/mutability issue to the user: http://www.win-vector.com/blog/2014/04/you-dont-n
898.
by jmount 12y ago
Original wasn't loading for me at the time, so I didn't feel good sharing what seemed like a link to a site momentarily experiencing trouble.
899.
by jmount 12y ago
Sounds like the publisher or studio was accused of being in turmoil back at the time (such as http://webcache.googleusercontent.com/search?q=cache:8Ex0abZ... ).
900.
Utomatic bias correction doesn’t fix omitted variable bias(win-vector.com)
1 points by jmount 12y ago|0 comments

More ›