achompas

58 ms·

601.
by achompas 14y ago
Net bailouts performed by the Fed: trillions http://www.bloomberg.com/news/2011-11-28/secret-fed-loans-un... . They gave banks and others at least $7.7 trillion and there is zero proof that it was paid back. It'd be nice if that article
602.
by achompas 14y ago
I'd agree, if finding such an algorithm was easy! :)
603.
by achompas 14y ago
The beauty of item response theory is that you can infer both the test's parameters and the student's abilities. You can use EM or Gibbs sampling to estimate those parameters simultaneously, so there's no need to limit yourself to test diff
604.
by achompas 14y ago
You've got the right idea--there's lots of recent literature treating the typical IRT problem as a Bayesian inference problem. Some of the references listed at the bottom of the article (especially the Kim one) provide great literature surv
605.
Using IRT Instead of Traditional Grading to Assess Students(knewton.com)
7 points by achompas 14y ago|6 comments
606.
by achompas 14y ago
Signed up. It looks like these guys really understand today's tech ecosystem. /serious
607.
by achompas 14y ago
He addressed it in today's Build & Analyze *(a podcast on 5by5 that he co-hosts with Dan Benjamin): http://5by5.tv/buildanalyze/80 Very excited to hear his thoughts.
608.
by achompas 14y ago
Fortunately Light Table will have support for packaging, so users can add any language they want.
609.
by achompas 14y ago
I'll chime in and say that, as one of Knewton's former interns (now a current employee), I had the opportunity to work on significant problems, learn as much as I wanted, and collaborate with some talented engineers. Knewton is an excellent
610.
by achompas 14y ago
That's arguably a result of Linux's small user base. Linux users tend to be tech-savvy but game-thirsty, so it's possible HIB managed to tap that market out first. My point: "stagnant Linux growth" might actually be "excellent Mac/Win growt
611.
by achompas 14y ago
The chart shows progress, but when you dig into the text above from the BLS website, there are some glaring metrics staring at you in the face. "Dig into" a two-paragraph blurb? Instead of looking at raw BLS statistics? Uh-oh... Why are
612.
by achompas 14y ago
Geez, we seem to have this conversation every week. If companies aren't being silly, business analysts != data scientist. You still need to be comfortable with various statistical distributions, learning algorithms, graphical models, geneti
613.
by achompas 14y ago
The fact that within the span of a few months Dan lost Let's Make Mistakes To be fair, Let's Make Mistakes wasn't "lost." Mike Monteiro actually spun it off into its own podcasting network: http://muleradio.net/
614.
by achompas 14y ago
Right. I hope this isn't Gruber's intent, given that he monetizes his brand (i.e. him ) by selling shirts and appearing at conferences. You can't really dictate the terms for interaction when you when you're such a public figure. If he is
615.
by achompas 14y ago
The $240 itself is irrelevant, the return on investment is what matters. I disagree ($240 might represent the entirety of their marketing budget for now, so the ROI on flyers could be moot), but regardless, parent did not mention ROI at a
616.
by achompas 14y ago
Really? They only spent $240. Would flyers for the neighborhood cost less than that?
617.
by achompas 14y ago
This is the best description I've read here so far. Scientific programming is ultimately a "niche" task, and there's a limited number of areas where this work happens (startup hubs, sure, but otherwise its research institutions, financial c
618.
by achompas 14y ago
I agreed with this post up to the point where you recommend a Masters in either statistics or economics. As a former econ student/researcher, I can say that you should not get a MS in economics or statistics if you want to do scientific co
619.
by achompas 14y ago
I really think you have it right: negotiate the release. Firing a month before OP's first year of equity vests is an incredibly low move. It's also suggested (and arguably correct) that OP might have legal recourse (the "option") if the com
620.
by achompas 14y ago
Hell, maybe learn how to code, while you're at it. It's not clear that OP DOESN'T know how to code. Also, OP wasn't employee #5--he or she was employee #1, and sacrificed a year of earnings as a result. You don't make that sacrifice, then
621.
by achompas 14y ago
Saw this a while ago and it's a cool thought experiment. The application fails, though, as some HNers have pointed out. It looks like Time covers have simply become darker, independent of the skin tones on each cover. Someone points out the
622.
by achompas 14y ago
Agreed 100%. Grad school is a time to explore all kinds of CS fields (or various topics within a field you're interested in). Right now I wish I had 28-hour days to mess around with GPU computing, compiler design, language theory, and rando
623.
by achompas 14y ago
Thanks for the tip--I held off on an Intel 160GB for $250 back in January, and now the same price gets me a 256GB!
624.
by achompas 14y ago
I agree wholeheartedly and would point people towards Chris Dixon's article on the topic.[0] Specifically, I think we're witnessing early-stage overinvestment. This has been anecdotally validated by Dixon and Fred Wilson, who state that c
625.
by achompas 14y ago
SSDs have dropped precipitously from about $2 a GB wholesale to about $0.78 or less a GB wholesale. Has this happened recently? I know retail != wholesale, but I haven't seen SSD prices drop at NewEgg for some time.
626.
by achompas 14y ago
If we're talking "strong signals," well, he already has MS on his resume. That's a pretty big signal.
627.
by achompas 14y ago
This comment isn't clear, but you make a few good points. Short-run profits are maximized by producing at the point where marginal benefit = marginal costs (I think you state as much: "If unit sold goes up by more than the lost per unit pro
628.
by achompas 14y ago
Thanks for the excellent summary. This line: I am good with mathematics but not the mathematics usually needed as a data scientist resonates with me. If you're thinking of data science, you're facing a loooooong road of coursework (scient
629.
by achompas 15y ago
This is the reason why I cannot be a "data scientist" Are you worried about this outcome at all? Do you see yourself playing an important role on a data team, one with less modeling responsibilities but more infrastructure/DB responsibili
630.
by achompas 15y ago
I don't notice an effort to expand the workforce by training or by recruitment of non-traditional workers, etc. "Code Year" for data scientists would actually need: -- "Code Previous Year," where everyone would need to kick ass on Bayesia

More ›