An Introduction to Fuzzy Logic

We taught a whole semester class on Fuzzy Logic and Fuzzy Systems and I think you covered almost 30 percent of the whole class in 3 minutes 48 seconds. This is amazing. Thank you.

[@alvinlepik5265]

I dare say the lectures are more involved than this short recap :P

[@TheSafe313]

Did your professor also have extreme vocal fry?

[@glowcarefoundation1998]

i agree to som extent but dis is just an overview

[@apollocreed8324]

Literally the best intro. Told about fuzzy logic basics in a matter of minutes. Great work mate keep it up

[@scotlandsimon]

Best intro to fuzzy logic video i found on KZbin, keep on man :)

[@lucidlagomorph5809]

wow if thats true then we are all lost. fuzzy logic still has to exist within a binary system it's about creating a programming language to simulate analog

[@raduduggan4667]

indeed, great video Tim. i think anyone that manages to breakdown something complex into small digestible bits deserves another subscriber!

[@ovanhoa4c-185]

Thank you so much, bro! I can immediately understand your video instead of reading the textbook for hours.

[@anonymuser302]

Rest in Peace Lotfi Zadeh...

[@MR-cs6vo]

Hours of lectures in 3 minutes nailed it

[@BeKindPlox]

This is a GREAT and compact explanation, thank you!

[@smithwill9952]

Well presenting the idea behind FUZZY. That's why lots of appreciation and subscription.

[@markradewagen5075]

In interpersonal communication, this would be like not knowing the details of someone's life, but interacting with them based on how a person typically likes to be treated. In machine learning, this seems kin to reinforcement learning.

[@majinbuux7]

Another underrated channel

[@ImtithalSaeed]

SIMPLE AND VERY POWERFUL EXPLANATION

[@Randomkitten7]

3:42 Satellites have attitudes. We need to keep them in control;) Nice explaination by the way

[@AdamNoizer]

Thank you. Helped massively with my project.

[@aminbazgir7873]

You did a GREAT JOB! Thanks for your explanation.

[@sureshr3882]

Wow..better example to understand a fuzzy logic ...

[@FockewulfAz]

This theory was introduced by Lotfi Aliasker Zadeh, Azerbaijani scientist.

[@sleepysundaymorning5034]

That’s crazy bro

[@fuadisgandarov1922]

@@sleepysundaymorning5034 As crazy as your useless attitude)

[@FockewulfAz]

@Elsevar Asadov sehf deyirsen. Reklam deyil. Haricler bilmelidir ki bizde savadli adamlar coxdur.

[@FockewulfAz]

@Elsevar Asadov duzdu savadli adam cixib gedir olkeden. Cunki onu qiymetlendirmirler ya de rusvet vermek mecbur eliyirler. Cahil qalmaq istemirsense Qurani oxu. Sen Quranin mentigi basa dusenden sonra, fergi olmayaraq professiyani oyrene bilersen. Islam adami cox deyisir. Men pesmen edirem ki namazi ushagligdan qilmiram ve Qurani o vaxt oxumadim. Bizimkiler olkeni o veziyyete saliblar cunki cahildiler, oxuyan deyiller...

[@fuadisgandarov1922]

@Elsevar Asadov Almanlara, İngilislere bele ikili standard tetbiq olunmur. Onları onsuzda hamı tanıyır. Bizim kimi balaca ölkelerin (ehali olaraq) dahileri sizin kimi insanlara göre unudulur çox teessuf ki. Men sizden özünüze hörmet etmeyinizi istemeyecem. Çünki görürem bacarmarsız). Heç olmasa burda bele gereksiz şeyler yazmayın.

[@kaayeni]

Quite an introduction. Very apt! Thank you

[@olinkirkland]

Would really like to see you make more videos, this was informative and easy to understand.

[@UNKNOWN-ms2tm]

BEST INTRO .... IN A VERY SIMPLE LANGUAGE ..... THANKS BUDDY

[@91722854]

so excited, as this is part of the syllabus of my module in my yr3 course

[@spsreenivaspadala122]

best ppt on fuzzy logic so far

[@rootsharp9946]

Wow. Sounds like Moriarty from Sherlock Holmes giving some lectures! Thank you! I'm a big fan of yours.

[@statisticsforbeginners6841]

Thankyou for the explanation .I am using the fuzzy logic in decision making

[@jayasankarvn]

Cool... Awesome examples and cool descriptions... keep it up

[@joshua50101]

Good and clear examples, nice representations of reality. Way to go dude, congrats are in order. Thanks a lot, keep it up :)

[@marcojustiniano7498]

Amazing thank you for breaking it down so clearly.

[@faboxbkn]

i am crying thank you!!

[@ashickakonjee8861]

Thanks for the nice examples and explanations.

[@balajidurairaj6838]

Excellent explanation 👌

[@madadibrahimov4845]

Rest in Peace Lotfi Zadeh

[@LD-wf2yt]

Fuzzy logic is the secret sauce ie. reason for bad or failing systems to continue to linger for much longer than necessary. Always stay aware of all levels of the Iceberg model of change.

[@oghuvwublessing705]

Thank you very much.

[@igniter1814]

Thank you for the info. The background music distracts and sometimes annoys a lot.

[@chintarisa]

Best intro...Nice sharing..

[@mohammadjavadnouri2623]

THANK YOU SOOOOOOOOO MUCH. That's the best intro

[@moc_yt]

Explained very well

[@anant8884]

brilliant Tim.

[@viraj_singh]

Really great explanation

[@mtwg9307]

Fussy logic actually applied for many years ago. Even before the term "fussy logic" was invented, people used this kind of digital process for closed loop feedback in wide range of application. Otherwise, just use 0 /1 for simple control is not realistic for application.

[@imranshaikh-tz5ik]

Excellent Explanation...Thanks

[@arpitbagri5010]

Very nice explanation ... Thank you sir

[@gizmoguyar]

Well, I'm still very confused. I understand the concept behind fuzzy logic. I just don't understand why people keep saying that it's so much better than PID controllers. In this example you have to measure the distance between the two cars and apply breaks based upon that measurement. Why is using fuzzy logic better than a PID controller? In this example a simple proportional controller would perform exactly the same, but the logic behind it is much simpler. The error would be the difference between the target distance and the measured distance. Multiple the error by some gain, and apply that much breaking power... In fact, it uses few calculations as well; one simple subtraction and one multiplication. Using fuzzy logic in this situation requires two full linear interpolations and two multiplications plus a final addition. Maybe this is just a poor example. Can someone give me a simple example where using a PID controller is much harder and/or much less accurate/desirable than fuzzy logic?

[@gizmoguyar]

Okay, to answer my own question, I believe I understand. A PID controller is simply one special case of a fuzzy logic controller. PID controllers either require a model of the system's behavior, or can only act upon measured factors. Fuzzy logic incoporates rules into the controller. For example, if the distance between the cars is decreasing very rapidly, apply more brakes. If the car in front is large apply more breaking power, if there is a car very close behind, apply less breaking power. All of these to a greater or lesser extent. A variable gain PID controller would be required to react in these situations.

[@marsag3118]

I have exactly the same doubt as you about fuzzy logic controllers. there is still something that does not completely convince me. Regarding the example you wrote in the answer above: in a PID controller isn't the fast / slow decreasing / increasing of the distance managed by the D part of the PID? so I don't see yet the breakthrough of using fuzzy controllers instead of PIDs, which, as you said, are much more straightforward, and their time / frequency response has been studied for decades.

[@JWEX500]

gizmoguyar so are you saying the main advantage of fuzzy is that it accepts multiple inputs whereas PID is strictly SISO?

[@gizmoguyar]

Jackie Wu I still don't believe that fuzzy logic has any advantage. PID controllers can definitely have multiple inputs and outputs. Tuning them becomes much more tricky, but it's the same system and thus no more tricky than fuzzy logic. and once again the PID beats fuzzy logic in speed and computational accuracy.

[@timarnett8555]

A PID controller is a 3-input (state error, accumulated error, and error rate), single-output controller that is linear. A Fuzzy Logic Controller can take in as many states, and give as many outputs, as desired and are universal approximators. Since some optimal control mapping for a particular system is more likely non-linear than linear, FLCs can more closely approximate this than a linear controller. To say that PIDs beat FLCs in performance is demonstrably false (see pretty much any paper comparing them). The benefit to using PIDs is because they are linear, we have analytical means for showing certain properties (stability, performance, etc). The non-linear behavior of (most) FLCs means it is more difficult to definitively show those properties.

[@orangeziggy348]

Tell me why it took until the 20th century by Lotfi Zadeh to finally apply/introduce fuzzy logic to mathematics, when this should have been already applied during Roman Times?

[@suttikarn043]

Fucking good explained! I´ve just spent 4 hours just to find out what Fuzzy logic means....... Thank you!

[@AhamedKabeer-wn1jb]

Thank you..Good Explanation..

[@mdalaminmahamud3775]

it literally made my perception clear about fuzzy logic :D (y)

[@shashankshetye7589]

Good explanation

[@AdnanAli-gd7wn]

Keep on bro! best intro

[@RaoMohini]

Analogue systems have continual variation of values. Digital systems have discrete steps of values. Fuzzy systems have discrete set of values. Right ?

[@shwetarani4370]

very nice explation! thank you

[@EhCloserLook]

Thanks for the explanation... but was it supposed to be "altitude" and not "attitude" there at the end of the presentation?

[@timarnett8555]

Nope. Attitude refers to the orientation of the object.

[@franklydoodle350]

I used that intro track for a cheesy 80's advertisement haha

[@ayoubjibouni1428]

thanks it's very clear easy and useful

[@adesatrio]

good explanation,thanks man

[@BM94849]

one simple correction if i may add, in the very last slide, you mention satellite attitude instead of altitude great video tho, this made fuzzy logic clears to me

[@timarnett8555]

Bogiva Mirdyanto No, attitude is correct. You can also control altitude for station keeping but the example I was giving (attitude) refers to the orientation. I.E. keeping the satellite pointed at a particular location.

[@BM94849]

+Tim Arnett ah so that's what you mean. aight then, nice explanation

[@rameshchandran]

Nice representations ... Thanks.

[@djedditarek682]

very good explanation

[@Math_Flicks]

Really helpful..

[@travelwithfarhad5827]

Thanks a lot for your contribitions

[@jasmeetdhanjal4407]

Love your voice

[@brotherschannel9810]

great video bruh

[@rameshchandran]

Nice representations ... Thanks...

[@cecitorresmx]

thanks, easy and useful :)

[@XDdemonslayer]

This is really no different than pressure transducers or encoders, they take a value at any range and assign a digital value so that it can be interpreted, and the resolution can essentially be as high as you like. It still follows the same rules of digital logic and Boolean algebra, I can guarantee you that a braking system in a car is more than just a 1 and a 0. Not quite that simple.

[@james77011]

Nice video ☺️

[@drajapaksa]

Great video... Can I have this slides ?

[@TheDradge]

It looks more like the digital (logic) levels of 0 and 1 are now just spread across analog values. Surely it's just a replacement of a binary logic (digital) system to an analog system.

[@warker6186]

so the fuzzy logic is all about numbers between 0 and 1. And traditional logic is just about 0 and 1 (just like binary). How close, how far & close, far. So the fuzzy logic should have a number range that in our case is close and far. So it may be close=.00 to .50 and far=.51 to 1 or 0.99

[@madrasnarayan]

Awesome 👍

[@25Aditya25]

how is it different from analog

[@Sketch1994]

So basically it's an analog system with a digital interpreter? Like a proximity sensor driving a potentiometer circuit, driving the brake solenoid at "analog" pressure...plus a screen!

[@cwk18]

That's exactly what I'm thinking, so basically fuzzy logic is analog version of digital system?

[@rameshchandran]

Nice representations ..

[@prightlesolutions490]

Thanks for this

[@skemsen]

I wondering: This makes no sense if one omits defining the terms. What is "cold", what is "warm", what is "close" what is "far"? And doesn't the mere concept of logic collapse when you deviate from it's own inherent property? Is the concept of logic even debatable from a philosophical perspective?

[@BobrLovr]

"This makes no sense if one omits defining the terms" Yeah thats kind of how every thing works lol

[@ni3cat]

Can someone explain me why we have a fuzzy logic definition. As it seems it is just bunch of if else statements and switch.

[@lucidlagomorph5809]

fuzzy logic is the process of a computer to learn that a horse can still be a horse even if its a white horse. I would have thought fuzzy logic as a concept has been around longer than the term was first coined. Grammar correctors would have to use fuzzy logic I would have thought. Nobody used the grammar correction schema and said ah this logical framework would be useful in other applications. I am surprised there isn't a patent dispute going on here

[@yasserd0153]

What is the difference between fuzzy logic and quantum computing?

[@gizmoguyar]

Quantum computing is the act of storing information as atomic spin. It has nothing to do with greater or lesser degrees. The pin is either down, up, or something else. The reason why quantum computers are being hailed as the next big thing, is because they'll be faster and smaller than our current transistor based processors due to their atomic nature. Fuzzy logic still requires analog to digital conversions and is still subject to errors due to resolution.

[@yasserd0153]

informative! thanks

[@ajithjayasundara423]

scorned certificate

[@arjetasahiti3707]

THANK YOU !!!

[@shakiulanam6569]

very pretty representation

[@amirawaheed3714]

Thanks alot 🙌

[@cloudberry7241]

Thank you!

[@ProfessionalTycoons]

thanks!

[@Mau365PP]

So... is it just proportional control with saturation? 🤷‍♂️

[@timarnett8555]

No. Proportional control is a type of linear control, although the saturation would indeed induce a nonlinearity at the threshold. Fuzzy controllers are almost always nonlinear (depending on their properties). If you have a Fuzzy Inference System (uses Fuzzy logic along with input-output inference rules) that gives a piecewise linear output, you could exactly match the output of a proportional controller with saturation with only a few membership functions. Since Fuzzy systems are (usually) universal approximators though, they can approximate much more complicated control surfaces, which is useful since an optimal control function is almost always nonlinear.

[@athiralbar2661]

Really sorry i am 5 years late but, if you were to read my comment, you mind giving me the sources of all the information used for the video bcs i am making a project report based on fuzzy logic and i barely have enough information about fuzzy based on any past studies, especially explaining it in simple terms

[@timarnett8555]

This is a good all-around resource: www.amazon.com/Fuzzy-Logic-Engineering-Applications-Third/dp/047074376X Of course there are several others but googling is probably good enough.

[@muhammadfaris1252]

very cool thx

[@devilzwishbone]

So in short Fuzzy Logic is an analogue version of classic logic which would be binary

[@shubhamchauhan4267]

Can you send me the link of the template?

[@brainlink_]

PER ITALIANI: Ho creato l'unica playlist su KZbin Italia sulla logica fuzzy! la trovate quì -> kzbin.info/aero/PL-tCoHPn6YlafYUe6bSnpZJe2MSaHer8h

[@hossainahamed8789]

sir Suppose, I am making a fuzzy rules using three input variables a, b ,c, and 1 output variable d. Each variable has three linguistic terms Low medium and high. Is it possible to write A B C D High Low Low Medium High Low Low High If so , can you give me a paper reference ? Thank you

[@timarnett8555]

I don't have a reference, but it is possible. However, you'll still need a way to choose one of the antecedents or combine them in order to resolve to a final crisp value. You could use a disjunctive rule like a probabilistic sum, or maybe a t conorm, to combine the two output membership functions. Or you could learn some sort of probabilistic method from training data. The end result however could also be achieved by having more granularity in your input and output membership functions.

[@timarnett8555]

Oops, instead of antecedent that should say consequent.

[@zoltanjosvai7008]

excellent

[@謝宗佑-u1m]

In short: It's relative.

[@SupremeST25]

Me trying to explain to my date how 5’ 6” constitutes “tall”:

[@prosoftwebindia]

Thanks!!! small clear..

[@theoriginalrichard]

Are you the same guy who does the casually explained videos??????

[@timarnett8555]

I'm not sure what you're referring to so....probably not.

[@kalpitoriLampung]

Mantap, salam dari indonesia

[@SafwanAlselwi]

Nice

[@Comezehere]

It’s not fuzzy. It’s a spectrum based on reality. For example if a car needs 2.1 seconds to brake and not hit the front car that’s because of speed and distance from the front car and the following cars engineering capability. Not to mention weather conditions. But if the car behind needs 7 seconds to avoid a collision (like a massive truck);then the variables change.

[@mos8916]

Mhmm, i do not think that the example of the car is the most suitable to explain why we need fuzzy logic. In fact, in that example an analog circuit would simply do the job.

[@glowcarefoundation1998]

You literally explained wht fuzzy is all about., nice one. You used 3min to explain what we are been taught in a semester.....sofunny

[@kiran7122]

which sensor is used to detect the distance between the car??

[@kaamadughaa3309]

doppler effect

[@TamimTasnimZiko]

Tnx