psb217

15 ms·

151.
by psb217 13y ago
While I can't take away your belief in a human right to something, I can take that something (given sufficient capacity for force). Belief in n/h rights can exist independently of whether or not the rights exist, in the same way t
152.
by psb217 13y ago
I said should and should not are useless "in the absence of a specific objective". But, as you suggest, they are useful in the context of an objective, e.g. I should not push old ladies into heavy traffic (implicitly: because I pr
153.
by psb217 13y ago
Basically in agreement with rayiner's sibling comment, I'd say morals don't exist in the physical/concrete sense that gravity exists, but they do exist in the cultural/behavioral sense that beauty exists. For them t
154.
by psb217 13y ago
"Legal rights don't exist, but natural and human rights do." Given that existence of legal rights can be proven empirically [1], whereas existence of natural and human rights is a philosophical quagmire yet to be resolved aft
155.
by psb217 13y ago
"Everyone is capable of autonomy." Thank you for so concisely and cogently resolving what may well have been philosophy's longest-standing and most ontologically profound line of inquiry. I'm confident that one day I wil
156.
by psb217 13y ago
That's not quite what "rotational symmetry" means in this context. A collection of random variables independently distributed according to identical "doubly-rectified" exponential distributions would also be symmetr
157.
by psb217 13y ago
Smaller drones with worse optics should be cheaply producible. With large numbers of such drones, lower altitude surveillance with greater redundancy of coverage would become affordable. The computational costs associated with patching toge
158.
by psb217 14y ago
Your second paragraph ignores the inference required to go from victim's blood on defendant's shoe to defendant's shoe was in physical proximity to victim's body. It also ignores the inference required to go from blood is measured as matchi
159.
by psb217 14y ago
Your choice of examples actually supports the statistical nature of the distinction between circumstantial and noncircumstantial evidence. That the blood on the defendant's shoe came from the victim's body is based on a probabilistic test.
160.
by psb217 14y ago
Or, you could just write: M = repmat(f(1:10), n, 1); where 'n' gives the number of rows you want in 'M' and 'f' is written in proper "Matlab" style (i.e. behaves reasonably when given an array as input). Or, to be more in the spiri
161.
by psb217 14y ago
In your final sentence, you appear to have mistyped "tangential" as "tendentious", which is an oddly sophisticated typo. Though, in this case, "ancillary benefit" may more accurately convey your (apparent) intended meaning than "tangential
162.
by psb217 14y ago
In addition to what bravura said, Geoff Hinton's group at UT has recently introduced a new approach to training neural networks that they call "dropout". You can read about it in their paper ( http://arxiv.org/pdf/1207.0580.pdf ) or watch H
163.
by psb217 14y ago
Tell that to Sebastian Thrun ( http://en.wikipedia.org/wiki/Sebastian_Thrun ), whose work on the Google driverless car project ( http://en.wikipedia.org/wiki/Google_driverless_car ) and current position as a Google Fellow owes so much to hi
164.
by psb217 14y ago
Conceptually, logistic regression is a simple neural network. A neural network with no hidden layers and a softmax output layer using cross-entropy loss is precisely equivalent to logistic regression. Though, if you were to extend your sugg
165.
by psb217 14y ago
The difference between "should" and "must" is significant. If I said: "people should not be hypocrites" and "people should be free to say what they want", my statements would not be incompatible or contradictory. In the same way, simultaneo
166.
by psb217 14y ago
If you look through chapter 12 of the most recent (online) version of ESL by Hastie et al, you can get a good idea of how to think about what the soft-margin SVM is doing. In particular, equation 12.25 on page 426 gives a formal mathematica
167.
by psb217 14y ago
Elements of Statistical Learning is Hastie's book. Between it and Bishop's book (i.e. Pattern Recognition and Machine Learning), I prefer Hastie's for clarity of exposition. In particular, I find that ESL better conveys the sort of intuitiv
168.
by psb217 14y ago
Your description of SVMs is a bit misleading. It matches with what are typically called "hard margin" SVMs, which do require linearly separable data. However, people talking about SVMs are typically talking about "soft margin" SVMs, which d
169.
by psb217 14y ago
When I said "at present", I was operating in the context of your previous post, i.e. assuming a universe with static space-time geometry. In this context, the present empty coldness of the universe is relevant and the past crowded hotness i
170.
by psb217 14y ago
Err, Einstein's static universe did not preclude the death of stars via exhaustion of their fuel. So, the entire universe would "heat up" to a uniform distribution (i.e. evolve according to the heat equation), with the future local heat bei
171.
by psb217 14y ago
Over the last few years, some people have started to work on a problem that has been described as "pure exploration" in multi-armed bandits [1,2]. The objective in this problem is, roughly speaking, to maximize the rate at which you become
172.
by psb217 14y ago
Referring to the number of nucleotides in BRCA1/2 as n, what happens if I develop a test that checks (1) if nucleotides 1->3n/4 of BRCA1/2 are present (2) if nucleotides n/4->n of BRCA1/2 are present (3) if the substrings in (1) and
173.
by psb217 14y ago
I hadn't seen that paper before; thanks for the reference. I read through it and saw that they were reweighting the sampled image patches with a Gaussian mask prior to learning, which explains how they got Gabor-like filters. The masking ef
174.
by psb217 14y ago
FYI, Gabor-like filters pop out from doing ICA (i.e. Independent Components Analysis), not PCA. While PCA looks for orthogonal vectors onto which the data's projection is normally-distributed (among other properties), ICA, roughly speaking,
175.
by psb217 14y ago
Yes, that.
176.
by psb217 14y ago
You seem to be ignoring the central role played by London financiers in the largest global destruction of "capital and jobs in the pursuit of notoriety, peer adulation and the odd $1bn exit" in the last 50 years. In other words, there's a p
177.
by psb217 14y ago
You may want to look at the paper: "P-packSVM: Parallel Primal grAdient desCent Kernel SVM" from ICDM 2009. It presents an extension of Pegasos to non-linear kernels. Evaluating the pairwise kernels <x_i,x_j> and continuously updati
178.
by psb217 15y ago
Precludes seems to be precisely opposite the intended meaning of your second sentence. Perhaps you were looking for precede, presage, or produce?
179.
by psb217 15y ago
Firearms are reifications of information, while hacking tools are information left incorporeal. Arguments founded on their equivalence are as devoid of meaning as those equating bits on a computer to cars on a lot. Consider the following qu
180.
by psb217 15y ago
Facebook only gets screwed if their own valuation doesn't increase enough to match the amount "overpaid" in response to Instagram's acute "increased valuation" ploy. Given current market conditions, particularly in terms of fawning over Fac

More ›