Kohonen self organizing maps(KSOFM) with algorithm and solved example
ฝัง
- เผยแพร่เมื่อ 11 ก.ย. 2024
- #neuralnetwork #softcomputing #machinelearning #algorithm #datamining
Neural networks | Self Organizing Maps | KSOFM | Solved Example
Introduction:1.1 Biological neurons, McCulloch and Pitts models of neuron, Types
of activation function, Network architectures, Knowledge representation, Hebb net
1.2 Learning processes: Supervised learning, Unsupervised learning and
Reinforcement learning
1.3 Learning Rules : Hebbian Learning Rule, Perceptron Learning Rule, Delta
Learning Rule, Widrow-Hoff Learning Rule, Correlation Learning Rule, WinnerTake-All Learning Rule
1.4 Applications and scope of Neural Networks
10
2
Supervised Learning Networks :
2.1 Perception Networks - continuous & discrete, Perceptron convergence theorem,
Adaline, Madaline, Method of steepest descent, - least mean square algorithm,
Linear & non-linear separable classes & Pattern classes,
2.2 Back Propagation Network,
2.3 Radial Basis Function Network.
12
3
Unsupervised learning network:
3.1 Fixed weights competitive nets,
3.2 Kohonen Self-organizing Feature Maps, Learning Vector Quantization,
3.3 Adaptive Resonance Theory - 1
06
4
Associative memory networks:
4.1 Introduction, Training algorithms for Pattern Association,
4.2 Auto-associative Memory Network, Hetero-associative Memory Network,
Bidirectional Associative Memory,
4.3 Discrete Hopfield Networks.
08
5
Fuzzy Logic:
5.1 Fuzzy Sets, Fuzzy Relations and Tolerance and Equivalence
5.2 Fuzzification and Defuzzification
5.3 Fuzzy Controllers
For Last Vector at (4,1) the value is not 0.425 it is 0.725.
But it was an great tutorial I have exams coming this really should help me. Thank you for your efforts keep helping us and future generations to come !!!!!
This guy is a life saver for my Neuro fuzzy exams!!! Thanku so so much!!!!!
Great video. Best explanation ever.continue with more videos please .👍👍👍
Very clear and precise. Keep it up bro.
Thank u so much for explaining such topics... your videos help me a lot.. many books are available but couldn't find such examples ...God bless u
0.75x 😂
What are you saying!! I watched it at 2x.
@@tppt3987 same
If you tried 0.5x you wouldn't have said that
thankyou for the video
No sé inglés y aún así entendí perfectamente. Excelente explicación👌
Thank you very much for this tutorial, friend :) It was very helpful.
well explained..thank you
The value for w41(new) that I get in the end is 0.725 for the 4th input vector.
Yes.... absolutely right !!
Wrong
Thank you again! I really do not know why people say you explain fast! I really learned very well, and you, you have backward button on your media player too.
There is less time in my life, I m gonna die soon.
@@hrb9877 There is forward button too! ;-)
@@nikronic Foraward button bafkward buttons dabane se acha video normal speed pe bana leta, I already running short of time no time to do forward and backward
I'm subscribed your channel for this video...
This video helped me in correct situation
you sir are a GENIUS, many thanks!!!
Many Thanks, but i think the very last figure (weight) = 0.725 Not 0.475 --> Calculation 0.45 + 0.5 (1*0.45) = 0.725
(1-0.45) not (1*0.45)
Yeah bro u r right
Yes bhai
Great explanation and demonstration. Big thanks. Please try to make a video on Back propagation with example like this one. Thanks once again!
Thanks man! much help
problem has been solved, many many thanks...........
This example can be found on the book "Fundamentals of Neural Networks" by Laurene Fausett (page 169). It is well explained there.
very nicely explained
Thank you so much! Very well explained!
Brooo thank you so much. Nice explanation👏
Good explanation sir
When would the updation of weights Stop? How do we know the number of epochs or iterations required ?
When there is no noticeable changes after weight updates
According to the video, when you finish calculating with the input vectors
por favor habla mas despacio. I like it!!! Good video, congratulations!!!!
Bro, mzze aa gye
i often joke about 'pajeets' but you guys dont get enough credit for how hard you study.
thanks for the video man.
Thank you so much! It helps a lot! 💜💜💜💜
When will we update learning rate? Not even once was it updated in the video
NICE VIDEO... please add a video to update the learning rate... show the final output i.e. output clusters
Good and clear explanations, Sir. Thank you.
Nice samja sab!
Very nice
Thank u so much
Excellent!
good explanation
thanks for the explanation
Thank you for your efforts
great video))
Thank you sir
Thanks for the video.
What is the convergence criterion?
Superb !
Thank you
really appreciate the hard work and tremendous job done :)
Is there a reference book you can provide for your tutorials ?
Nice video but don't we update the learning rate?
Superrr bro
Bhai thoda slow kr le... BTW great work
Right
Kyaa aap ki train ki late hore the kya bro
Good
Very good tutoring my question is, how could clustering data from a matrix, like an image.
nice video, can you please update the video of CPN
Please add more videos as my exams are very soon
Bhai dhanyawad 🙏pass kardega tu bhai
you are genius
thank you so much that was very useful for me
by the way, results of w new is false
This question can be stupid but I have to ask. So If we have 3 values in inputs (instead of four , like [1,0,1]) how we calculate then ?
You started directly with an algorithm, why didn't explained the SOM first
So what happens after this? I update the learning rate and keep going i suppose...
But how long? What is the condition for stopping? I have to set limit to how small the learning rate can be? (for example set it to 10^-4) Or am missing something big here?
itna ghai me kaha ja rahe the bhai??
itna speed tha ki man hi ni kiya dekhne ka
😘😘💞
@btech tutorial plz tell me the source of the question from where did you took the question ?
Hey pillai guys....
Please, how did you get the values for the initialized weights?
Small Random values have been considered.For one Cluster in increasing order, For the other in decreasing.
@@btechtutorialByNishantMittal so, I can choose any random number that is less than 1
Yes u can.
where did the value of alpha come from ?
Ç Plz provide video of Linear Programming Problem using graphic and DuAl SIMPLEX method with examples
bro you have written wrong formula for eucledian distance at 2:35, it should (xi-wij)^2
However both give you the same result....
How identity alpha?
i adore your speed, which is not found in the other Indians, no offence.
Step 4 learning rate α kb update krenge apne explain bhi nho ki 4th step
bhai bhot speed se kya sab kuch aap ne 😀
Please teach the linear algebra instead of the arithmetic :p
Tereko slow motion mei sune kya, kitna fast bhagega....yaha pe slow learner aate hai isliye toh colg mei samaj nai aata....LD candidtates please like me
How value is ...W41 =0.475 ??
Reduce speed to 0.75, use your pen and paper to follow along
Can i get notes of this subject ?
Nice . But too fast . Slow study wins the race.
How to visualize that code
flash who?
When do we use step4..
Learning rate a=0.5a
After one complete iteration.
When do you update the value of alpha?
in tunning paramters, where you try different learning rates
Y so hurry??
Terko etna kaha bhag kar jaana tha. Etna fast tutorial banaya.
what about step 4???? updating learning rate???
Learning rate is multiplied by 0.5 for every epoch
Plss explain faster bro...Bore ho gae itna slow me
why this much speed?? absolutely not needed.
r=?
teach slowly bro pls
ussain bolt is slower than u
Is that Sachin Tendulkar's voice??😂😂😂
haha
thank you , hope you drink after :D
If possible use Hindi language for explaining
you are very fast i can’t understand
already 1.25x, lol
Say slow bro not catching an express..
Bruh don't mean to be rude but its ok to catch ur breath once in a while instead of rushing it.... Everything else was good nice explanation
please speak little slow
O.75x is the key to watch !
Huh! and I was playing this at 1.5x
@@AbhinavKumar-qn1dt It ain't EDM bruh
Drop the beat
@@AbhinavKumar-qn1dt hahahha
HAHAHAHHA ARE YOU IN A RUSH OR WHAT
I think it's his way of teaching
Thank u so much