That's my first time to review your teaching! Very thanks with you and proud for the indians, yours electronic education is the best! I am come from HK!
this only look larger because you are compressin a message of 0's and 1's imagine your A was the zero and the B was the 1 if you use this methode on this example you only need 4 bit to code the leter A, and you would need 5 bits for example to code it (25 leters or something>16)
the sequence ABAB would require 4 bits instead of 5*4 bits that would require without compression i was having the same question, it looks like it doesnt compress but it actually does
When you don't have a very long sequence like in this case, there is no redundancy so the compression actually produces even longer sequences. Imagine a data stream of thousands of KB's, you would have a lot more redundancy and this algorithm is gonna take out that redundancy to produce compressed streams.
I was trying to understand this topic for about half an hour, but I could not understand anything, all understood from this video within 8 minutes..Thank you so much...
what happens if the data does not end at ...1010 but instead ends ...101 ? "101" has been seen before, but there are no more symbols left... i suspect we can add extra padding bit (or more) so that we create a new symbol?
@@LalitVashishtha so during the first phase, we read 101 at the end and since this sequence was seen before, we ignore it? This makes more sense, also thank you SO MUCH for answering!
Bonjour, Je veux savoir ça si possible, 3- La compression doit normalement générer un mot de code de taille réduite. expliquez moi, pourquoi l'algorithme de Lempel-Ziv n'a pas pu réduire la longueur du message de source dans l'exemple: AABABBBABAABABBBABBABB 4- Aidez moi par un autre exemple dont la compression effectuée permet de réduire la taille du message de source.
Dictionary location always starts with 1 upto the number of codewords. LZ algo is fixed length encoding scheme. Thesefore, number of zeroes are appended in LHS in dictionary locatilon. It depends on the base of the number system. For binary numbers willbe 0 and 1 only, in ternary numbers will be 0, 1, 2; in quaternary numbers will be 0, 1, 3, 4 and so on.
Location always starts with 1. Here info sequence is in binary and there is 8 info seq. We need dictionary location of prefix, so we need 1 to 7 locations.Since LZ algo is fixed length encoding and If we take location of 8th info seq it will take bits that is 1000 and this location will never be used. So we shall take locations from 1 to 7. 7 is the largest no here which will be written as 111 in binary therefore 1 will be written as 001. 2 as 010. 3 as 011 and so on.
Dictionary log starts with 1 to no of info seq. Since it is fixed length encoding technique all the locations will har same number of digits. Here 1 is written as 001, 2 as 010, 3 as 011, 4 as woo Tec.
your explanation is wrong, you are saying you see 8 codewords so there will be 3 bit dictionary location. for 010100110110, I divide, 0,1,01,00,11,011,0 and then according toyour explanation I will need 3 bit location entries, since i have 7 types of symbols to code. but for the above mentioned problem i will need actually 2 bit dictionary location.very misleading explanation
I don't know which book you are following, but if you go through "Information Theory, Coding and Cryptography" by Prof. Ranjan Bose, IIT Delhi; you will get the same concept as I discussed here.
Finally one clear explanation in English, thanks !
This explanation is very good and I clear Lempel-Ziv coding. Thanks sir.
Though your English in unclear you explanation is clear as water. Hats off bro. Kudos
Great Teacher. ! clear and calm while teaching complex theories
thank you sir! that helped me a lot in my finals!
That's my first time to review your teaching! Very thanks with you and proud for the indians, yours electronic education is the best! I am come from HK!
at last found it . good english .good explanation . nice shirt . thank you
Thanks
thank you teacher
Thanks to you, I understood this part that I didn't listen to in class
You teach better than my IIT professor
We can learn anywhere but exposure we get iit is incredible so do not compare
Btw i am iitian
Thank You For making it easy a day before exam sir!!
thanh you for making it easy a day before exam sir!!!
Great professor. Good explanation ♥️
This really has saved me , I got exam tomorrow...Thanks Sir
Sir yours lectures are good and understand able for students
Thank you for it
#keep supporting
Best video on LZW so far,,
Thank you sir , loved the expression at 4:56
thank you so much sir, a clear explanation
Isn't the result of the algorithm much larger than the original string? How does it work for compression? Thank you!
this only look larger because you are compressin a message of 0's and 1's
imagine your A was the zero and the B was the 1 if you use this methode on this example you only need 4 bit to code the leter A, and you would need 5 bits for example to code it (25 leters or something>16)
the sequence ABAB would require 4 bits instead of 5*4 bits that would require without compression
i was having the same question, it looks like it doesnt compress but it actually does
When you don't have a very long sequence like in this case, there is no redundancy so the compression actually produces even longer sequences. Imagine a data stream of thousands of KB's, you would have a lot more redundancy and this algorithm is gonna take out that redundancy to produce compressed streams.
I was trying to understand this topic for about half an hour, but I could not understand anything, all understood from this video within 8 minutes..Thank you so much...
Well said..
Easy to understand when u explain
Can u also explain
Deflate and PNG compression techniques
thanku so much sir...how easily u explain this..
thank you sir for your way to explain this theorem
Bappu tu pass hoagaya ... baghwan tumhe salamat rakhe
Clear and Perfect Presentation !
Best explanation thankyou so much sir
Neat and clean explanation.
Thanks from Ecuador !!! :)
Great explanation sir
Amazing explanation sir!
thank you so much.
Thnx I got 5 marks question from this topic in my end exam.
Lalit ji, tum bahut mast kaam karta hai...
Aren't You Supposed to Assume that 0,1 are already stored in the Dictionary 🤔
same doubt
Thanks for your support
thx, my english is bad but i could understand you
How data compressed here ? Encoded bits are more than original bits why?
Bahot sahi explanation Sir...!!👍
tommorrow is our semester exams & your videos are really helpful. Thanks a lot lalit sir
Thankyou very helpful
Beautiful explanation...
Nice Explanation!
its nice. but show the full board at least for 3 seconds please
Thank you so much, very helpful.
the lsb name is innovation symbol
Flawlessly explained
Best Teacher!!!
Sir, you are just awesome.
Thank you sir what about a matlab function (used to calcule lempel-Ziv parmetere EEG )?
Perfect explanation
thanks
Thank you so much sir
what happens if the data does not end at ...1010 but instead ends ...101 ? "101" has been seen before, but there are no more symbols left... i suspect we can add extra padding bit (or more) so that we create a new symbol?
No, no padding is done in this situation. We have to write the codeword of 101 which has already been found.
@@LalitVashishtha so during the first phase, we read 101 at the end and since this sequence was seen before, we ignore it? This makes more sense, also thank you SO MUCH for answering!
Bonjour,
Je veux savoir ça si possible,
3- La compression doit normalement générer un mot de code de taille réduite. expliquez moi, pourquoi l'algorithme de Lempel-Ziv n'a pas pu réduire la longueur du message de source dans l'exemple:
AABABBBABAABABBBABBABB
4- Aidez moi par un autre exemple dont la compression effectuée permet de réduire la taille du message de source.
Howw u find dictionary locationn iss ir already given orr that we have to findd first??? I not getting thiss could anyone please clear this out
Memory location always starts with 1 and increases 1 in next step and so on...
Your awesome sir
Sir you explained only the encoding part. Do explain the decoding of the same. The rest was simple to understand and very helpful.
I will try to upload decoding video by this week
I will try to upload it by today itself
Great video sir!
Can u explain wht is dictionary locatiom
thanku so much sir
beautiful explanation
many thanks, sir !
thankyou🙇♂
Sir i could bot understood dictionary location
Remains is excellent
Dictionary location always starts with 1 upto the number of codewords. LZ algo is fixed length encoding scheme. Thesefore, number of zeroes are appended in LHS in dictionary locatilon. It depends on the base of the number system. For binary numbers willbe 0 and 1 only, in ternary numbers will be 0, 1, 2; in quaternary numbers will be 0, 1, 3, 4 and so on.
in above comment make on correction for quaternary daya numbers will be 0,1,2 and 3
Nice
i m confused how to get RLE for 101010 could u solve
What do you mean by RLE here?
@@naoromunda2731 run length encoding
(11,10)
Since 10 is repeated 3 times
Very nice sir
Thank uh so much Sir
Good example
Excellent!
How to write dictionary location ?tell me sir
Location always starts with 1. Here info sequence is in binary and there is 8 info seq. We need dictionary location of prefix, so we need 1 to 7 locations.Since LZ algo is fixed length encoding and If we take location of 8th info seq it will take bits that is 1000 and this location will never be used. So we shall take locations from 1 to 7. 7 is the largest no here which will be written as 111 in binary therefore 1 will be written as 001. 2 as 010. 3 as 011 and so on.
Dictionary log starts with 1 to no of info seq. Since it is fixed length encoding technique all the locations will har same number of digits. Here 1 is written as 001, 2 as 010, 3 as 011, 4 as woo Tec.
Thank u sir
0 00101110010100101 iska phase batao 8 sequence ayenge Kya? Or 9
0, 00, 1, 01, 11, 001, 010, 0101
Sry but.... Humare clg k sir ne 9 sequences likh KR diye!!!!.
Like: 0 1 00 01 011 10 010 100 101
He will be using something else, you may go throgh Information Theory< Coding and Cryptography by Prof Ranjan Bose IIT Delhi
Tq so much sir
Cannot solve AAABBCCDDABBA using lzw
thank you !
.thank you.Sir can you please tell me tge text book for reference
Information theory, coding and cryptography by Ranjan Bose TMH publication
Thank You Sir
good one
Arigato gozaimasu !!!!!!
Thanks
Thanks slot
Effort👍👏
your explanation is wrong, you are saying you see 8 codewords so there will be 3 bit dictionary location. for 010100110110, I divide, 0,1,01,00,11,011,0 and then according toyour explanation I will need 3 bit location entries, since i have 7 types of symbols to code. but for the above mentioned problem i will need actually 2 bit dictionary location.very misleading explanation
I don't know which book you are following, but if you go through "Information Theory, Coding and Cryptography" by Prof. Ranjan Bose, IIT Delhi; you will get the same concept as I discussed here.
link me your concept or explain it if possible.
@@LalitVashishtha 0,1,01,00,11,011,0 can be encoded as 00(0), 00(1), 01(1), 01(0), 10(1), 11(1), 01
Put some link or name of the book you are following
Tq
Tqq sir
tq
we won cricket match, muhahaha
You are not solving any doubts of any student so why you telling us to write...
thankyou sir