tgflynn

87 ms·

931.
by tgflynn 14y ago
Anyone who has actually spent much time working on Windows vs. Linux with large files and probably more importantly large numbers of files is aware of how much more of pain this is on Windows. Theoretical issues aside Windows FS implementa
932.
by tgflynn 14y ago
C# is essentially Windows only and Windows is not a good platform for heavy data analysis. Windows has many problems but one of the worst is poor file system performance. On Windows you waste a lot of time moving large files around when yo
933.
by tgflynn 14y ago
If it takes as much effort to write the unit tests as it does to implement the solution then the problem was trivial to begin with, but that's probably true of 90% of all programming work. It might be more useful for harder problems but the
934.
by tgflynn 14y ago
It will be a good idea when you can actually get paid for implementing contracts.
935.
by tgflynn 14y ago
This is great news. As skeptical as I am about the prospects for actually simulating the human brain in the near future the benefits that are likely to arise from devoting more resources to understanding it are enormous, both in terms of
936.
by tgflynn 14y ago
Not sure if this would work but have you tried quoting the values in the generated CSV ?
937.
by tgflynn 14y ago
It sounds like they're saying 300 neurons is too hard so let's do 10^11 instead. Science usually proceeds somewhat incrementally from easier problems to harder problems and I'm not seeing the foundation here. Weren't simple genomes sequence
938.
by tgflynn 14y ago
Do you have a reference ? Do these worms exhibit non-trivial behavior or is it at the level of say reflex attraction to light sources ?
939.
by tgflynn 14y ago
I would feel less skeptical about the prospects for simulating the human brain if I had heard that someone had succeeded in simulating an infinitely simpler nervous system, say that of a spider, for instance, and obtained spider-like behavi
940.
by tgflynn 14y ago
As far as the C++98 standard library goes I think vectors are by far the most useful construct it provides because they're like arrays but manage memory for you. Lot's of people prefer printf to cout/cin. Iterators in my opinion are a bad i
941.
by tgflynn 14y ago
I sort of understand how you feel and sort of not. I used to think I was a great programmer, now I'm not sure, because I've seen evidence for and against. But in the end how do you ever know ? For example I was once asked to develop a JPEG2
942.
by tgflynn 14y ago
SEEKING WORK - Remote (Based in Rochester NY) Specialties : Computer Vision (Including OpenCV) Machine Learning C++ Development Project based or hourly ($100 per hour). No web development. Contact: tgflynn@cogniception.com
943.
by tgflynn 14y ago
About 50X.
944.
by tgflynn 14y ago
It seems like if a 1 megaton impact were occurring on Earth every 15 years people would have noticed more than seems to be the case. You would expect more than one such event per century to occur on land and I think the vast majority of th
945.
by tgflynn 14y ago
Most of the ionizing radiation from an H bomb is produced by nuclear fusion or from fission from the trigger and decay products. An asteroid impact wouldn't produce any of that. The only ionizing radiation it might produce are some x-rays
946.
by tgflynn 14y ago
The undeniable observation is that certain people spent a long time trying to do perception with a certain set of tools and found it to be very hard. That does not necessarily imply that perception is intrinsically so much harder than logic
947.
by tgflynn 14y ago
Well what I mean by "hard" is that I can train a neural network to do a reasonably good job of recognizing say handwritten characters given some amount of training data. My attempts to train similar networks to learn the rules of chess (jus
948.
by tgflynn 14y ago
It's an interesting observation and certainly something one needs to be aware of in thinking about intelligent systems but I'm not convinced that the notion that perception is intrinsically harder than logic is quite true. I think the diff
949.
by tgflynn 14y ago
I think my argument was mistaken. The CNF form I was thinking of involves adding unknown variables so it doesn't actually allow you to compute the function in one step.
950.
by tgflynn 14y ago
If you restrict the depth of the circuit, you can in principle represent any function, but you need a really wide (exponentially wide) circuit. Are you sure it's exponential ? If you look at binary functions (ie. boolean circuits) any suc
951.
by tgflynn 14y ago
The claim that you don't need more than 1 hidden layer is based on a mathematical theorem that says roughly that a sufficiently large 3 layer net exists that can fit any sufficiently smooth function. One has to look in detail at things lik
952.
by tgflynn 14y ago
The USPS address recognition technology (at least as of 2 years ago when I was working on it) is not human-level. A fraction of the images cannot be resolved and are still sent to human keyers at rec sites. This fraction has been decreasi
953.
by tgflynn 14y ago
If you're a skilled developer you can easily make that much freelancing on elance.com.
954.
by tgflynn 14y ago
I left a corporate job a couple of years ago to pursue technology development in areas that I think will be making a huge impact across all industries in the near future, notably computer vision and artificial intelligence. One of my bigges
955.
by tgflynn 14y ago
I haven't done much with geographic data myself but I know Google has an API for plotting that kind of data : http://gislounge.com/how-to-import-data-make-maps-google-fus... Beyond that R is a very powerful tool for data analysis. I woul
956.
by tgflynn 14y ago
With enough data you could build a model to predict, with some level of accuracy, the total number of tickets that will end up being sold for a given event as well as the likely time and geographical distribution of those sales. In order to
957.
by tgflynn 14y ago
It's true that my formulation isn't completely specified but the basic idea still works because if you have a decision function for an NP problem you can find a solution (assuming at least one exists) with one call to D per bit of the solut
958.
by tgflynn 14y ago
In the case of this problem the set of possible paths is finite so it can, in principle, be sorted in order of increasing cost. The smallest cost delta between any 2 paths in the sorted list is the lowest cost edge in the graph (this probab
959.
by tgflynn 14y ago
That's exactly what floating point numbers are: representations of real number intervals as finite length bit strings. The number of bits simply determines the granularity of the representation (ie. the width of the intervals).
960.
by tgflynn 14y ago
No revolution intended. This is well known. The revolution would be finding a polynomial time D.

More ›