redcalx

19 ms·

151.
by redcalx 10y ago
See https://en.wikipedia.org/wiki/Jevons_paradox
152.
by redcalx 10y ago
I believe your comment is missing a /s tag.
153.
by redcalx 10y ago
The heat map is the baseline data model; from that it's possible to derive simpler models for simpler use cases. However, providing access to the baseline model will likely be useful for some use cases. i.e. if you want something more
154.
by redcalx 10y ago
> To claim they should simply not return any data if they do not have an accuracy level to your arbitrary standards would make the service useless Ideally they'd offer version of the service that gives a probability 'heat map&#
155.
by redcalx 10y ago
Yeh good compromise. The ideal solution would be for their API to not report a precise location if it doesn't know one. I guess they have no simple method of stating a region rather than a point, and if they did the consumer of the API
156.
by redcalx 10y ago
I like how they 'fixed' it by moving the default to the centre of a lake. So you still have the waste of police (or whoever) driving out to some location before they figure out what's going on.
157.
by redcalx 10y ago
> Convolution is not a matrix multiplication in the current implementation I figure there's a code re-organisation task since propagating node activations through a layer of weights is essentially a matrix multiplication (fully conn
158.
by redcalx 10y ago
1) Expected cost of failure = probability * cost given failure occurs. Many people will fail to correctly estimate expected cost when the probability is low but the cost number is high. 2) In this case the probability estimate was also wro
159.
by redcalx 10y ago
Brilliant, I've been looking for projects like this. I'm currently working through a couple of RBM C# projects but will add this to my list of reference code. Top tip: If you use the matrix and vector classes in math.net then you
160.
by redcalx 10y ago
What happened to OpenCL? https://en.wikipedia.org/wiki/OpenCL
161.
by redcalx 10y ago
Cut USB cable; splice new device into cable. Or, open mouse/keyboard case, wire device into USB bus connections.
162.
by redcalx 10y ago
There's also an annual allowance (in the UK anyway), so you can sell enough to make 11k (approx.) profit a year before capital gains tax kicks in.
163.
SharpNEAT – Evolution of Neural Networks(sharpneat.sourceforge.net)
1 points by redcalx 10y ago|0 comments
164.
by redcalx 10y ago
> You'll find that all sorts of everyday objects are optimized for tall people Cool, I should move there (6'6"). Then again if enough people did so their average height would increase even further.
165.
by redcalx 10y ago
That would make sense - an evolutionary adaptation to lower light levels.
166.
Dutch men confirmed as world's tallest(bbc.co.uk)
89 points by redcalx 10y ago|146 comments
167.
by redcalx 10y ago
It would be cool if you could use this to find people with similar taste, i.e. mentally and geographically proximate.
168.
by redcalx 10y ago
Perhaps a more serious point is that a single power failure (in one data center) can have such a wide impact.
169.
by redcalx 10y ago
As an aside, I've found BT's DNS to be very flaky; sometimes down (or unreachable at least) for many hours at a time. I switched to google's and never had any problems since.
170.
by redcalx 10y ago
> threatened to go nuclear and change the proof-of-work protocol, No miner with any significant amount of hash power would install such a release.
171.
by redcalx 10y ago
Incidentally, your (The Guardian) android app is very slow to load, it's annoying enough that I uninstalled it.
172.
by redcalx 10y ago
> proof of weak fairness Interest piqued. Do you have a good link for an intro to this?
173.
by redcalx 10y ago
> quadruple-speed single precision execution and double-speed double precision FYI Xeon Phi supports AVX512, i.e. 512 bit SIMD registers (8 doubles or 16 floats). In the consumer CPU world there is AVX2 which has 256 bit registers (4 dou
174.
by redcalx 10y ago
Domestic solar isn't metered at all. The power can go out to your neighbours on the same substation, but how much of it does so is not metered - payments are based on an assumption that 50% of the power generated is 'exported&#x27
175.
by redcalx 10y ago
They likely use half precision (i.e. 16 bit) floats in some of the computing stages. They've also discussed using even fewer bits of precision in some of their research output. When training and following an error gradient you kind of
176.
by redcalx 10y ago
I would suggest that you consider 'voting with your feet' so to speak. You are, after all, ultimately paying for that wasted energy with you backblaze fee.
177.
by redcalx 10y ago
See the other comment: there is a lot of data that isn't accessed for years at a time, so you're essentially spinning disks for years to save a few seconds of spin up time. I think he's right - energy is too cheap for this to
178.
by redcalx 10y ago
Shannon entropy is a measure of how much new information each symbol in a communication channel gives us. In the case of a PRNG, if we know the seed then each symbol tells us nothing because we can generate the exact sequence from the seed,
179.
by redcalx 10y ago
Wondering if data centres typically power down a drive after some period of no accesses? Seems like you could save a lot of power and gain some HD life that way - so long as a drive isn't frequently spinning up and down. Seems like the
180.
by redcalx 10y ago
Another option to consider: Craig is being pressured to make this claim to flush out the real Satoshi.

More ›