murbard2

67 ms·

631.
by murbard2 11y ago
Does an up to date IntelliJ offer the same full set of functionality as PyCharm? As in, is PyCharm just IntelliJ minus some stuff, or does it have features IntelliJ doesn't have?
632.
by murbard2 11y ago
When I had my wisdom teeth removed as a kid, I was absolutely terrified about the metaphysical consequneces of general anesthesia, so I know where you're coming from... Here are a few thoughts - Think of the universe not as evolving th
633.
by murbard2 11y ago
Big difference between letting kids be free to do things and not coddling them vs. using authority to push them to do things they are uncomfortable with or afraid of.
634.
by murbard2 11y ago
Even assuming his fears as a kid were blown up, they are still real and the dad's attitude towards them is extremely cavalier. I had similar fears as a kid, and I grew out of them. Being pushed never did anything to alleviate those fea
635.
by murbard2 11y ago
Yes it is well written and the style is light and refreshing, but am I the only one who think this is horrific child abuse?
636.
by murbard2 11y ago
Excel can be a great prototyping tool. I'm a statistician/programmer, and yet, I sometimes like to explore some ideas quickly in excel. It has a reactivity that many programming languages like... it's very nice to change a pa
637.
by murbard2 11y ago
Cute, but an implementation of the mental poker protocol would be far more valuable.
638.
by murbard2 11y ago
The Tesla PowerWall comes with a ten year warranty. Assuming a charge and discharge every day, that's 3560 cycles... but this kind of batteries can probably handle 500 full cycles at best. What gives?
639.
by murbard2 11y ago
A lot of the advice you will get is going to be in the spirit of the American way of teaching various scientific disciplines. I would decribe it as a hands on approach, that gets you started really fast and doing exciting things. This is tr
640.
by murbard2 11y ago
Not just "stimulus", but active exploration of the environment, communicating with and learning from other people, etc.
641.
by murbard2 11y ago
Even if you cannot experience death, but you can experience a horrible near miss with massive brain damage. For quantum suicide to work, you'd want to be completely unconscious when it happens, so that you never observe any injury.
642.
by murbard2 12y ago
Sorry but no matter how good he was, it's ridiculous to say he was the entire talent. I'm in a good position to know: I worked for several years within the group he was a part of (right after he left) and there were tons of terrif
643.
by murbard2 12y ago
1.32^(1/10)-1
644.
by murbard2 12y ago
You're missing my point. In many real world problems, it is cheap to compute the gradient. Thus, black box optimization methods which can use gradient information are inherently valuable, and it is surprising that they do not have a tr
645.
by murbard2 12y ago
It's a little strange that they do not have a track that gives gradient information, given that it is often a real world possibility. Also, this basically allows unlimited time between eval... So this becomes a contest about - coming
646.
by murbard2 12y ago
The advantage of brute force generation is that you can eliminate a lot of solutions extremely fast by trying random inputs.
647.
by murbard2 12y ago
This hits the sweet spot. It's more flexible than a DSL like Caffe which gets in the way as soon as you stray even a little bit, without requiring as much boilerplate as a generic framework like Theano. I have a question: When training
648.
Darts, Dice, and Coins: Sampling from a Discrete Distribution (2011)(keithschwarz.com)
15 points by murbard2 12y ago|1 comments
649.
by murbard2 12y ago
For a second, I thought this was going to be about adding smells to the experience... which would probably be a very bad idea, as nauseating experiences condition strong smell aversions. I wonder if a nose pincher would decrease motion sick
650.
by murbard2 12y ago
Not so much the computing power as a few tricks: - proper initialization of weights (should have been obvious, but no one really saw it) - rectified linear units and dropout
651.
by murbard2 12y ago
Yes, it's an expression. The algorithm is evaluating that expression. However, algorithms also evaluate algorithms, and a better example would have been an actual algorithm that needs to be interpreted.
652.
by murbard2 12y ago
"Taxpayers are spending a lot of money subsidizing not people who won't work, but industries that don't pay their workers a living wage" The hidden assumption in that statement is that those industries would pay their wo
653.
by murbard2 12y ago
Using a probabilistic program to generate and interpret a program is an interesting idea, but I think you're missing the main problem. How do you sample from the set of programs that produce the correct (or approximately correct) outpu
654.
by murbard2 12y ago
It's a very powerful statistical technique, yes, but I doubt it will be enough for AI. The problem is that you need to sample efficiently from the posterior distribution, and for anything AI related, MCMC isn't going to cut it. Le
655.
by murbard2 12y ago
Yes, you can specify extremely powerful statistical models in only a few lines of code using probabilistic programming. However, at this point, unless you design your program in a very specific way and use a lot of tricks, your sampler is v
656.
by murbard2 12y ago
Satisfying a complicated constraint is not much easier than sampling from a "thin" manifold.
657.
Xero, a SaaS company, went public without any sales(businessinsider.com)
1 points by murbard2 12y ago|0 comments
658.
by murbard2 12y ago
4. "Howard Wiseman and his team demonstrated that this wave function collapse is a real effect." No, they didn't. An alternate (and simpler) explanation being that the experimenter and the particle are entangled.
659.
by murbard2 12y ago
Heat conduction is very poor; you're practically in a vacuum, so you can only lose heat by radiating it. On earth, you can dump your heat into the atmosphere or the oceans, which are both gigantic heat sinks. The entire surface of the
660.
by murbard2 12y ago
You could also not count homework towards the GPA. I studied in France where the system is a bit different. We had homework, and they were graded so we could get feedback, but the idea that a homework grade would count towards your GPA woul

More ›