zodiac

54 ms·

541.
by zodiac 10y ago
I think so too - are you thinking of building such a thing? I noticed you had some open data science ipython notebooks on GitHub.
542.
by zodiac 10y ago
> There is nothing personal about my question. As I see it, this guy has questionable credentials to get such a prestigious job, so I ask what has he done. If that's all you did, I would say that's not a personal attack. But ho
543.
by zodiac 10y ago
It applies the unsupervised learning technique of https://en.wikipedia.org/wiki/Autoencoder As for why it's important and not just another encoder like mpeg4, see my response at https://news.ycombinator
544.
by zodiac 10y ago
> But what's the point? Autoencoders don't really have much practical use now, but that's not the point of them. The point is we really want to figure out how to do unsupervised learning well and autoencoders are one of th
545.
by zodiac 10y ago
> No, stockfish uses minimax with pruning... Minimax actually needs to look at all possible branches all the way down. No, I assure you that stockfish definitely does not look "all the way down (ie to states where the game is over)&
546.
by zodiac 10y ago
In that case, stockfish doesn't use minimax either right, because it does pruning as well
547.
by zodiac 10y ago
AG does use minimax in the same sense that for example stockfish does (bounded depth and with pruning)
548.
by zodiac 10y ago
So 3d models (like https://nizuvault.files.wordpress.com/2015/01/gavr-ship-rend... )?
549.
by zodiac 10y ago
I think they can. > In our pilot, the income will be unconditional; we’re going to give it to participants for the duration of the study, no matter what. People will be able to volunteer, work, not work, move to another country—anything.
550.
by zodiac 10y ago
I'm not a "BI fan" and I'm curious what they think too, but I think it could be done. I think it depends on the kind of emergency. For not-really-critical ones (eg Joe's car broke down), you could just do nothing -
551.
by zodiac 10y ago
> Apparently CS authors no longer bother with expanding acronyms the first time they are used. What does CS stand for? :p But yeah, I might expect BLEU to be unexpanded in an NLP paper maybe, but not here
552.
by zodiac 10y ago
> you target a market that people organically search for in the app store, with zero (or totally inadequate) options currently available) How did you go about finding such markets?
553.
by zodiac 10y ago
Did you use any regularization? Did you train until the training accuracy converged? I feel like the network has enough capacity to overfit 180 training samples
554.
by zodiac 10y ago
I'm very surprised that this worked with the small amount of training data he provided! (he "provided the algorithm ten images containing hand written digits..." - if we generously estimate that each image has 10 training ima
555.
by zodiac 10y ago
> we'll have our first batch of papers coming out soon are you batching papers instead of releasing them when they're done?
556.
by zodiac 10y ago
> What stopped Patey and Yokoyama from proving Ramsey's Theorem For Pairs/Triples by saying "for any pair which satisfies some relation X, there exists another pair which also satisfies relation X"? Because it's
557.
by zodiac 10y ago
Yeah, but it sounds like the extreme end of ad-hoc.
558.
by zodiac 10y ago
> like somebody tried lots of things and one seemed to work
559.
by zodiac 10y ago
Not all banking is rent-seeking; some of it creates some value
560.
by zodiac 10y ago
I agree with this. I think it will be a while before we get a decent theory, though, and it's not because no one is trying, but it seems too hard to me. For instance, the way I see it, we don't even know very well how the low-dime
561.
by zodiac 10y ago
> If you're choosing a specific model of math, you need to specify that in your statement. I mean, in my most recent statement I quantified over all consistent models... > It will still be computable within that model, just not p
562.
by zodiac 10y ago
Sure, but no matter what model of math we choose to work in, I can come up with some n such that I can say "...hence, BB(n) cannot be computed", and you cannot reply "BB(n) can be computed; it's just a finite number. Wha
563.
by zodiac 10y ago
You can't just say that things are either adhoc or not adhoc - it's a spectrum. On one end we will have engineers coming up with architectures by consulting a RNG and then testing out those architectures. On the other end we will
564.
by zodiac 10y ago
Yeah, sure, if we agree to do math in the system ZFC+Con(ZFC) we can prove that Yedidia's machine does not halt. But then we can just construct another machine (call it M', a n'-state turing machine) that searches for contrad
565.
by zodiac 10y ago
The type hinting is optional, so if you want "late type binding" you can just omit the type hints.
566.
by zodiac 10y ago
If you write down a value for BB 7910 but can't prove that it is correct, I wouldn't say that you "computed it".
567.
by zodiac 10y ago
If you're refering to Yedidia and Aaronson's machine, the machine searches for counterexamples in ZFC. We believe, but cannot prove, that ZFC is consistent. Also, we can prove that assuming ZFC is consistent, ZFC cannot prove that
568.
by zodiac 10y ago
You deliberately ignored the other part of his explanation. They didn't try out using an external tape just because they were inspired by Turing machines, but because previous RNN neural net architectures had a problem in that memory c
569.
by zodiac 10y ago
Firstly, (1) and (2) are not rules of thumbs for building nice architectures, but basic properties about supervised learning and what kind of problems it can solved. Even if it were 5 rules of thumb instead of 3, "design an architectur
570.
by zodiac 10y ago
I can't speak to the RL methods used here, but wrt to the architecture they designed, I think it might seem less ad-hoc if you know that generally speaking new neural net architectures work well when they 1) can do some useful computat

More ›