law

16 ms·

121.
by law 15y ago
Randomness can be a curse, but can also be a blessing when introduced as in the random subspace methods. This again abstracts to understanding your business needs and whether the results encountered make sense given the features' [absence o
122.
by law 15y ago
I have a good amount of experience in natural language processing and machine learning, and I don't think offering an API that provides easy access to the algorithms is the right solution. The major algorithms in text classification aren't
123.
by law 15y ago
Invalidating patents with prior art isn't as easy as you might think. There are two ways to do this: you can invalidate on novelty grounds (that is, the invention was anticipated by a single piece of prior art) or non-obviousness grounds (a
124.
by law 15y ago
Unfortunately, Blackboards' patents preclude this from becoming a reality. For the next 15 years, Blackboard has an exclusive right to practice this "virtual classroom" business method.[1] Their patent pledge[2] promises that they will not
125.
by law 15y ago
The House Judiciary committee. It hasn't been voted on yet.
126.
by law 15y ago
For all those who are interested, http://judiciary.house.gov/hearings/hear_11162011.html is the link to the hearing's webcast, which began at 10 a.m. EST.
127.
by law 15y ago
There is no benefit to society and "the Progress of Science and useful Arts" if somebody has just an idea (that anybody else could have too - maybe about a cool device, a cool algorithm, a cool user interface), publishes this idea and gets
128.
by law 15y ago
It's a fine line to walk. On the one hand, community-vetted code is a spectacular idea for the core algorithms, but on the other, overly-restrictive licenses (like [L]GPL) effectively preclude the maximum utility being derived from them.
129.
by law 15y ago
You're correct to identify the point of libraries like Weka and Mahout, which are both written in Java, as providing a solid framework for interaction between and among your program and other algorithms. However, Java isn't the right soluti
130.
by law 15y ago
Patents are designed to provide an incentive for companies to invest in heavy research and to publish the results, while being guaranteed a limited monopoly on the invention in order to monetize it. (Or at least that's what they should be
131.
by law 15y ago
Honestly, frameworks like Mahout and Weka have their place, and that's typically for exploratory data analysis. My belief is that for large-scale, extremely intensive machine learning, your best bet is to implement algorithms tailored to th
132.
by law 15y ago
Yeah, that's exactly the type of 'noise' I was referring to. Improper histological staining, poor imaging skills, etc. are all things that the human mind can immediately discover. The human mind is smart enough to not rely on the assumption
133.
by law 15y ago
So I am actually far less concerned about a computer doing my job very well, which is actually looking at a piece of tissue on a slide and making a tumor versus not-tumor distinction. This is very hard to do and I think will continue to be
134.
by law 15y ago
I don't think a reasonable person would ever claim the technological equivalent of a right of rescission for data contributed to a community-collaborated document. Rather, people are taking exception to unconscionable clauses in the terms o
135.
by law 15y ago
Yes, I'm saying that inferring the information from the chat transcripts is unforeseeable to the users generating those transcripts. It's obviously foreseeable that if you change your relationship status to "engaged," you might begin to see
136.
by law 15y ago
It's wonderful to see the EU upgrade its 1995 Directive to give users even more control over how data they voluntarily contribute is used. From the source, "This is why in our view, EU law should require that consumers give their explicit c
137.
by law 15y ago
Difference in definitions: I read "approximate solution" as an "approximate route," as in the node choices are approximated (potentially leading to a wrong final node, or losing packets at a dead end). Instead, finding approximately the sho
138.
by law 15y ago
That's the point, though: I don't know if you can find an 'approximate' solution to decentralized routing, since you need precision. Do you have any peer-reviewed articles evidencing these convincing results? I'd be extremely interested in
139.
by law 15y ago
After reading the article and skimming some posts on their subreddit, I think the idea generally concerns the capabilities of consumer electronics to 'replicate' the Internet in a completely decentralized fashion. By doing so, there's no ce
140.
by law 15y ago
Software still isn't eligible for patent protection, as I've said several times before (look at my comment history). Copyright protection and trade secret protection are available for software, while patent protection is available for "bu
141.
by law 15y ago
If I'm not mistaken, that was just a recitation that replaced the regular Thursday class. It was one of the TAs covering that stuff briefly. All three topics were covered by Tom Mitchell in previous classes.
142.
by law 15y ago
Actually, this has been actively researched since ICs started gaining widespread usage in the 1970s! Even before that there were plenty of journal papers produced that deal with the basics of ML and AI. It wasn't until the 1990s that comput
143.
by law 15y ago
There's this enormous focus on 'web scale' technologies. This focus necessarily invokes visualizing and making sense of terabytes and eventually even petabytes of data; conventional approaches would take thousands or millions of man hours t
144.
by law 15y ago
It's actually fine to compare an ensemble method (using weak base learners) to a single instance strong learner. In this way, you compare the benefits of combining the weak learners with the benefits of using a single classifier. I see wher
145.
by law 15y ago
We don't have "software patents," since software (as code) isn't patentable subject matter. Instead, we have "business method patents," which cover the systems and methods used by the software in conjunction with an operating environment th
146.
by law 15y ago
Ah, thank you, I wasn't even aware. It nevertheless should be something that's an opt-in feature, rather than an opt-out annoyance.
147.
by law 15y ago
...which means it should go to referendum, rather than be subjected to executive decree.
148.
by law 15y ago
Sorry if you mistook my post for me taking a side. I wasn't. I was just pointing out a peculiarity I noticed in the article. I'm actually for increased bike routes, as that will decrease the number of people injured in bicycle v. motor ve
149.
by law 15y ago
I'm well aware of the technical distinction between "libre" and "available." I should have said that Cassidy's belief is more in line with "libre," rather than gratuitous. A viable argument could be made that decreasing parking spaces inf
150.
by law 15y ago
The distinction is that it's private parking for $30 USD per hour. Municipal parking is provided by the taxpayers, just as the bike lanes are. They're switching out municipal parking for bike lanes when the bike lanes are underutilized.

More ›