superfx

Main paper here:

http://research.microsoft.com/apps/pubs/?id=211941

That's the entire point of my post. I think you may not have scrolled past the first picture, or perhaps not even clicked the link.

I did say "nearly" ;-)

For me the biggest shock about visiting Paris the first time was that it was a "real city", and not a boutique town where everything is perfect and manicured. Although, what was most impressive about it was the fact that a real city could be so nearly perfect and manicured (at least the rich parts).

Except that they're not just providing a language, but a back-end for cloud computing.

Well they've discovered that many stresses can induce this response, including physical stress. So it's not just acidity.

In many ways what's most interesting about this from a biological standpoint is that it explains why stem cells have been found all over the body in previous experiments. This "stress-induced" pluripotency suggests a general mechanism that the body uses to deal with stress. In effect cells are made with the intrinsic capability to respond to injury, by simply turning into stem cells.

I find Workflowy to be excellent.

My first impression reading the title was "don't waste your time going there".

I took this class in person at Stanford, and it's hands down one of the best and most useful courses offered in CS/ML/Stats. If the online course is anything like the real course then this is a must.

Rob Fergus is joining Facebook too:

https://plus.google.com/103199870243230279182/posts/JaqgU1sX...

I believe your reasoning is incorrect. If we model the number of fires by a poisson distribution (very reasonable), a single sample would correspond to the mode of the distribution. What you would like is an estimate of the mean, i.e. lambda. It so happens that the mode of a poisson distribution is floor(lambda) or ceiling(lambda)-1, and thus it is an underestimate of the mean. In other words, fires are likely to be less frequent than what a single sample suggests.

Thanks for the comment. I just fixed it.

I think it has been useful to me. For one, I'm more motivated to work because if I slack off, I'll be staring at an empty indicator all week long which really upsets me. I suppose it's just a psychological trick that may or may not work for everyone, but it does for me.

The second thing this helps with is project-based tracking. That is super useful, because I had no idea before how much time I'm really dedicating to my different projects. Of course, my initial assumptions have to make sense. I.e. if I decide to spend 25% of my time on a high-risk project and 35% of my time on a more attainable project, those targets have to make sense. But assuming they do, this system helps me match my targets.

I suppose the next step would be to regress those indicators on some external output, like number of papers published, but I think coming up with a sensible metric would be very difficult.

I'm the author of the original post. I use Dynamic in two ways. The more obvious way is just to create interactive visualizations that I can easily manipulate, for example changing the time granularity or the set of projects displayed. The second use is the link with my calendaring program. I use Dynamic to set up a timer that calls a DLL (through .NET/Link) every few minutes to programmatically fetch the day from Outlook. That way my dashboard is constantly up to date without me having to do anything, and so I can just glance at it whenever I want.

Since we're comparing notes, here's the Mathematica version:

Reverse[SortBy[Tally[StringSplit[#]], #[[2]] &]][[;; 10, 1]] &

I have specific requirements, particularly numbers 3 and 4. AFAIK none of the ones that show up on google can do those, in particular 3.

If I recall in the demo, they kept on referring to procedural generation, and so perhaps they have a few tricks up their sleeve for generating content (trees being an obvious example) that would be varied yet take up little memory.

It's a great graphic, but I'll nitpick. There's an important distinction that's left out. During all the stages leading to the Ph.D., all the area covered was just learning, taking in what other people have done. The very last little bump on the other hand actually added new area, so it's a very different thing, you're discovering new knowledge as opposed to just learning about what's already known. I suppose the graphic does capture that in a way but it's subtle.