Soft Computing

Hi! This is actually a case by case answer as a biologist will look for changes in appearance and productivity. Take the example of a mutant plant that is statistically 20% shorter. The decrease in biomass production is noticeable. On the other hand, if a mutant has an enzyme 30% less active than that of wild type, the difference can be very easily compensated by regulatory changes. Hence, the mutant may behave exactly like a wild type. This is one of the reasons why we typically consider a twofold threshold for gene expression changes even if smaller variations are statistically significant.

There are three general solutions to the non-duplicate random number problem: If you want a few numbers from a large range then pick one and reject it if it is a duplicate. If the range is large, then this won't cause too many repeated attempts. This is what you mention above. If you want a lot of numbers from a small range, then set out all the numbers in an array and shuffle the array. The Fisher-Yates algorithm is standard for array shuffling. Take the random numbers in sequence from the shuffled array. If you want a lot of numbers from a large range then use an appropriately sized encryption algorithm. E.g. for 64 bit numbers use DES and encrypt 0, 1, 2, 3, ... in sequence. The output is guaranteed unique because encryption is reversible. The Hasty Pudding Cipher can be set for any convenient range of numbers. Another answer from my side.

Ubuntu with more then 8 version will work perfectly and it is user frindly some what like window .. Suse or fedora is also option

In addition to 'Google Scholar', you can also consult with 'Microsoft Academic Search'. For security journals, please visit http://academic.research.microsoft.com/RankList?entitytype=4&topDomainID=2&subDomainID=2&last=0&start=1&end=100 For security conferences, please visit http://academic.research.microsoft.com/RankList?entitytype=3&topDomainID=2&subDomainID=2&last=0&start=1&end=100

Dear Sotirios Goudos, I would like to read the paper. Kindly mail me on aarti.gehani@nirmauni.ac.in

The values are positive real numbers (continuous) and the estimated values produced by the inference algorithm of constructed Bayesian network.

@Prof. Baruah ... All logic operators/operations in fuzzy logic return a fuzzy number, not a crisp 0/1; this fuzzy number reflects the spread of not-necessarily-uniquely-assignable evidence across the entire spread of possible responses. Ask yourself: have you *ever* said "That statement is somewhat true, somewhat false, and somewhat I-don't-know"? Or have you always responded to every question with a strict Yes/No, black/white? If the former is totally unknown to you, and the latter is the way you deal with every query, then we have nothing more to discuss -- you see the world and everything in it as being strictly dichotomous. I wish you a good life, and good luck. If you recognize the 'kinda-yes/kinda-no' phenomenon, then we're dealing with a fuzzy situation, one which is ideally suited to linguistic reasoning wherein the intersection between a thing and its opposite CAN in fact be non-empty ... because some subset of the evidence is, by its very nature, ambiguous. If you insist that, were everything to be perfectly understandable and perfectly calculable and perfectly assignable, THEN we could use a strictly Bayesian, probabilistic approach, well ... maybe. Those are some pretty strict restrictions and limitations, and I'll bet that they can only be asymptotically approached; if that's the case then we'll never see that ideal state, and no matter how close we get to it, there will always be that tiny little bit that we'll still be able to investigate or describe using fuzzy logic. However, *I* don't think we'll *ever* get all that far along that curve; *I* think that non-probabilistic uncertainty will *always* be a major factor in our reasoning scenarios, particularly those we engage in during our social interactions, i.e., when we talk to one another, which is when we make a HUGE number of our actual decisions .... yes? Don't *you* make many of your decisions while you're talking with someone? Can you then accurately and honestly claim that, while you are engaging in this discourse, you are also engaging in a purely internal, completely objective, entirely probabilistic reasoning process? *I* certainly can't make that claim ... particularly not when, for example, I'm talking with my wife. <waiting for the laughter to subside> Bottomline: If you do NOT see everything as being completely black-or-white,if the concept of ambiguity, conceptual-and-or-moral-and-or-ethical grayness has a referent to your life, has meaning to you, is something you have actually encountered and had to make decisions anyway, then ... you have already used fuzzy logic, you simply didn't know it. Regards to all, Doug

This question was originally asked by David Wolpert. In his worked he derived the No Free Lunch Theorem. This states that if you have no prior assumption on the nature of the problem we cannot expect an algorithm to be superior to the other. To my knowledge this theorem has not been proven incorrect. A good reference to introduce you to this theorem is the book by Duda, Hart and Stork "Pattern Classification" and the work by Wolpert. Hope this helps

Wilcoxon-Mann-Whitney test is good approach to find out the similarity between the data vectors. other is Kolmogorov-Smirnov test ks.test(x, y, ..., alternative = c("two.sided", "less", "greater"), exact = NULL)

In matlab you can use the "plotconfusion" function.

Bittorrent uses a tic-tac algorithm. If you upload more you will download more. Increase the limits of your upload bandwidth you will get more. In addition you can distribute the download over more than machine. This technique needs a small programing skills but it can work

You can stil hybridize. In AIS the IS system building is similar like evolutionary algorithm, where u can apply the strategy of SA , so that the SA can run in parallel one subunit of AIS.

@Xuanlong Ma: Thanks

You might check these papers: 1) S.-H. Cha. Comprehensive survey on distance/similarity measures between probability density functions. Inter- national Journal of Mathematical Models and Methods in Applied Sciences, 2) A. Strehl, J. Ghosh, and R. Mooney. Impact of similarity measures on web-page clustering. In Workshop on Artificial Intelligence for Web Search (AAAI 2000), pages 58–64, 2000.

For those that want the reference for Kohonen's original book it is: Kohonen, T. "Self Organizing Maps" 3rd Ed. Springer (2001) ISBN 3540679219

Net Logo Agent technology Language is the best

Google already remembers passwords that can be stored across computers. For more information see http://support.google.com/chrome/bin/answer.py?hl=en-GB&answer=165139 Here is some documentation to get you started http://developer.chrome.com/extensions/getstarted.html

thank you Bhavani....I already have gone through, but my specific query is regarding dominance based rough set's properties.Whether I can apply the same of rough sets in case of dominance based rough sets. Anyways, thank you for the help and I appreciate your initiative.Thanks again.

thanks alot for your response

topological structure of fuzzy soft sets

It is nice to have such materials for researchers. Prof. D. P. Sharma

As I have mentioned in several other discussions on Science 2.0 there are many financial, and Proprietary Rights factors that need to be thought of, the interface is not the issue, per-se. Is a mere comment on a discussion group, tantamount to publishing? Or is it more like a mini-seminar that is open to reading in the future? When we put our research onto a cloud, who owns it? the scientist that did the research, or the owner of the Cloud. This gets even cloudier when we look at the fact that the whole idea of the cloud is that you don't know where it is, or even how many computers have your research on them. I have been trying for years to find a way to finance a Grid Computer that I could use for research, People are naturally much more willing to offer me a single virtual computer than they are to offer, a 16 node cluster, or even the power, and storage location for me to install such a computer in. The idea that I could use the grid computer to finance it's own expansion, seems doomed to power issues and cooling politics. I never realized that air conditioning could be so political! It turns out that computer clock cycles are easy, but location is difficult. When you design a world-wide cloud, one of the problems is, how does the installation pay for its Internet access, real-estate, power and cooling? Does the cloud have a utility component where it charges for its use, or does it have free use, and if so how does the individual cluster pay for its own power and cooling? Who pays the rent on the server room? It makes sense to use a cloud concept only if there are sufficient funds to pay the costs of the support for the clock cycles involved. Boing Boing Sync, has had a cloud going for a considerable time, where the clock cycles were donated. The size of the clusters tends to be small, I understand, because they are mostly donated by individuals. The Grid on the other hand, I assume has a pay as you go structure. (Does it accept Pay-Pal?) CERN, the supercomputer place in Zurich, seems quite interested in the Network as an interface to global super-computing, so interested that we can blame them for the World-Wide-Web. Is this because they want to sell more clock cycles when they are not running the simulations for physics, or because they want to borrow more cycles from the web, when they are running the simulations. IBM, and NEC have massive installations that they might offer for clouds when they are superannuated, but in the meantime are being used for environmental simulations etc. What happens when such a super computer installation is sold, and the new owner can strip its hard drives of massive amounts of data? I once met a man in a garage. that had a mainframe, and was transferring data from the old main frame format to new media. The question was, why was he working out of a garage? if his work was valid, he should have had a lab at the University, but the potential was that he was stealing the data from the old media, and shipping it offshore, to allow scientists in his own country access to it, without the original owners consent. When we have our computer cloud based in Romania, where there are more criminal hackers than anywhere else in the world, can we be assured of any privacy at all? or will some Chinese Scientist publish from our results before we can? I have to say that I would prefer a Grid Solution where you have at least some idea of who is party to your data, rather than a cloud solution where you are hoping anonymity of the server, means anonymity of your data.