Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Fuzzy Logic

en.wikipedia.org

Post reply on HN
Print view View original

31–40 of 43 posts

Re: Fuzzy Logic

#31

Post by lonelappde »

tr352 wrote: post #20
elcomet wrote: post #17

Earlier quoted context omitted.

Yeah but in the real world it doesn't matter does it? you cannot compute the probability A and B for a reason: they might be related (what if they're mutually exclusive for instance? ) So how does fuzzy logic deals with this?

Fuzzy logic deals with this in the exact same way as classical logic. Note that in classical logic, the truth value of "A and B" and "A or B" is also a function of the truth values of A and of B. When using fuzzy logic you have to make a choice as to which function you use. Typically these functions generalise the classical logic ones in the sense that they behave like the classical ones when using 0 and 1. These fun…

That's great in toy theory, like the formal real number system.

How useful is it for modeling real world problem?

Re: Fuzzy Logic

#32

Post by aasasd »

badsectoracula wrote: post #30
pjbk wrote: post #14

Back in the 90s when neural networks and fuzzy logic were the rage and were going to solve all problems of human mankind (déjà vu?), I stumbled upon this nice course put together by Texas A&M in the form of an "ebook" (aka Win3.1 application). Lofti Zadeh himself was a contributor. I looked it up in my bookmarks and I cannot believe that the site is still up: http://faculty.petra.ac.id/resmana/private/fuzzy I was ver…

Out of curiosity i downloaded and downloaded the tutorial and it looks to be more of a hyperlinked multimedia(ish) application, similar to something you'd see made in Hypercard (which makes sense since it seems to be made in ToolBook which essentially was a Hypercard clone for Windows 3.1) and less of something like a PDF, something you'd see in an Amazon Kindle or something you'd see as an EPUB (which is what i thin…

I've seen stuff like that back in the early- to mid-2000s. It seems, of course, that most of such ‘multimedia programs’ were educational packages for kids.

But most of all, this reminds me of PowerPoint presentations, which iirc were sometimes similarly made for use at home and with educational content. Also olde .HLP and .CHM ‘books’ were employed in this vein.

Re: Fuzzy Logic

#34

Post by jacques_chester »

lonelappde wrote: post #31
tr352 wrote: post #20

Earlier quoted context omitted.

Fuzzy logic deals with this in the exact same way as classical logic. Note that in classical logic, the truth value of "A and B" and "A or B" is also a function of the truth values of A and of B. When using fuzzy logic you have to make a choice as to which function you use. Typically these functions generalise the classical logic ones in the sense that they behave like the classical ones when using 0 and 1. These fun…

That's great in toy theory, like the formal real number system. How useful is it for modeling real world problem?

Very useful. Many problems are more easily represented as fuzzy sets or fuzzy relations than in other terms. It's particularly good at encoding linguistic variables, such as "very fast", "too cold", "accelerate hard" and so on in a way that it can smoothly overlap them.

It's also useful for encoding uncertainties that are not yet mutually exclusive. There are other logics too (eg Dempster-Shafer evidence theory), often grouped together as "monotone measures".

Re: Fuzzy Logic

#35

Post by jacques_chester »

tlarkworthy wrote: post #11
tehsauce wrote: post #8

"Fuzzy logic is a form of many-valued logic in which the truth values of variables may be any real number between 0 and 1 both inclusive. It is employed to handle the concept of partial truth, where the truth value may range between completely true and completely false." Can someone explain how this is different than bayesian statistics?

Bayesian statistics has to follow Bayes rule. There is a fairly specific framework behind the numbers in Bayes (i.e. probability). Fuzzy logic is looser. You could say Bayesian statistics is a subset of Fuzzy logic. Given how informal people have to be in Bayesian statistics to come up with reasonable priors (e.g. uniform), and how well it works by just guessing reasonable values, it could be argued that the power of…

> You could say Bayesian statistics is a subset of Fuzzy logic.

Arguably not. Mathematicians have teased out differences between different many-valued logics and systems. A critical one between probability and fuzziness is that probability includes the axiom of the excluded middle and fuzziness does not. In probability the values of mutually exclusive events must sum to 1.0, in fuzziness they need not, because it doesn't require events to be mutually exclusive in the sense that probability requires.

Re: Fuzzy Logic

#36

Post by jacques_chester »

pjbk wrote: post #14

Back in the 90s when neural networks and fuzzy logic were the rage and were going to solve all problems of human mankind (déjà vu?), I stumbled upon this nice course put together by Texas A&M in the form of an "ebook" (aka Win3.1 application). Lofti Zadeh himself was a contributor. I looked it up in my bookmarks and I cannot believe that the site is still up: http://faculty.petra.ac.id/resmana/private/fuzzy I was ver…

Also available as a free textbook: Fuzzy Control by Passino & Yurkovich - http://www2.ece.ohio-state.edu/~passino/FCbook.pdf

I recently read Fuzzy Logic with Engineering Applications by Ross. Helpful and interesting, though he does tend to occasionally spend pages performing every step of a calculation.

Re: Fuzzy Logic

#37

Post by jacques_chester »

sacado2 wrote: post #22

Since we're talking about non-classical logics, I also like modal logic, which deals with uncertainty too, but in a very different way. Here, facts / assertions are not associated with a numeric value, but with a modality, a symbol, meaning "it is possible that F is true", "it is not necessary that F is true", "F is usually true", "F was true yesterday", or "I know I don't know whether F is true or false". https://en…

The truth value being 0.8 can be mapped to one or more fuzzy sets. That's what you're typically doing. Saying "I don't know" is still possible, because you can define a fuzzy set representing "I don't know". But now you can say how strongly you don't know.

An overlapping alternative is Dempster-Shafer evidence theory, which has "belief functions" and rules on how to combine them.

Re: Fuzzy Logic

#38

Post by jacques_chester »

kinow wrote: post #5
mooze wrote: post #4

A great application of this in real life is rice cookers.

I think washings machines, some digital cameras, and also a train station system in Japan were some other famous applications (from what I can remember).

Automatic gearboxes, ABS, cruise control, some plant controls. Fuzziness saw strongest uptake in Japan, so any control system developed there has a decent chance of having a fuzzy system at its core.

Re: Fuzzy Logic

#39

Post by pjbk »

badsectoracula wrote: post #30
pjbk wrote: post #14

Back in the 90s when neural networks and fuzzy logic were the rage and were going to solve all problems of human mankind (déjà vu?), I stumbled upon this nice course put together by Texas A&M in the form of an "ebook" (aka Win3.1 application). Lofti Zadeh himself was a contributor. I looked it up in my bookmarks and I cannot believe that the site is still up: http://faculty.petra.ac.id/resmana/private/fuzzy I was ver…

Out of curiosity i downloaded and downloaded the tutorial and it looks to be more of a hyperlinked multimedia(ish) application, similar to something you'd see made in Hypercard (which makes sense since it seems to be made in ToolBook which essentially was a Hypercard clone for Windows 3.1) and less of something like a PDF, something you'd see in an Amazon Kindle or something you'd see as an EPUB (which is what i thin…

Pre-powerpoint app presentations were popular back in the days. In the 80s I recall there was a huge industry of educational presentations for 8-bit Commodore and Atari machines. Some good, some bad, and some very bad. There were a few that were interactive and "hyperlinked". In particular I learned programming Atari ANTIC display lists and players & missiles with one that allowed you to modify things in an editor/debugger in realtime so you could immediately see the changes in the graphics, and also checked your correct responses at the end in the self-evaluation section. Pretty awesome. I wonder if they used an external software to program that or they just did everything themselves, as it was usually done then.

I just downloaded DosBox, installed Win3.1 and the software. I am having a blast down memory lane. It's impressive how your brain can retrieve memories so well with only minimal hints after more than 20 years... the persistence of memory!

Re: Fuzzy Logic

#40

Post by badsectoracula »

aasasd wrote: post #32
badsectoracula wrote: post #30

Earlier quoted context omitted.

Out of curiosity i downloaded and downloaded the tutorial and it looks to be more of a hyperlinked multimedia(ish) application, similar to something you'd see made in Hypercard (which makes sense since it seems to be made in ToolBook which essentially was a Hypercard clone for Windows 3.1) and less of something like a PDF, something you'd see in an Amazon Kindle or something you'd see as an EPUB (which is what i thin…

I've seen stuff like that back in the early- to mid-2000s. It seems, of course, that most of such ‘multimedia programs’ were educational packages for kids. But most of all, this reminds me of PowerPoint presentations, which iirc were sometimes similarly made for use at home and with educational content. Also olde .HLP and .CHM ‘books’ were employed in this vein.

I guess you could make something similar with PowerPoint and "jump to slide" actions and VBA macros, but i haven't seen any HLP or CHM (even though CHM could technically do it via JavaScript) that work like that.

Still, what i found interesting was mainly the way this was presented in bite-sized pieces (via the pages that couldn't arbitrarily extend via scrolling), combined with clear graphs/pictures and interactivity (not just hyperlinking, but also end-of-chapter quizzes and in a couple of cases showing values changing live as you move your mouse over them).

Technically simple stuff, of course, and certainly done before (i mean, ToolBook was made in 1989 and is still around, so chances are someone is using it :-P), but not something i see often, which seems to be a shame. I had only heard the name "fuzzy logic" before, but after reading that presentation (just the fundamentals) i got a decent idea about what they are and it was easy to follow the tutorial. In comparison the same URL has a bunch of PDFs, including a text-only tutorial that seemed to describe more or less the same stuff. However i could only glaze over before closing it since i just couldn't focus on it. The interactive stuff is just so much more attractive :-P

Post reply on HN