Тёмный

AN ELEMENTARY PROOF OF BERTRAND'S POSTULATE! Special #SoMe1 

Подписаться
Просмотров 8 тыс.
% 257

I love when a deep result in mathematics is provable only with elementary techniques, like basic knowledge of combinatorics and arithmetic.
In this video I will present the queen of this proofs, namely the Erdős' proof of the Bertrand's postulate, which states that it is always possible to find a prime number between a positive integere n and 2n.
In general such kind of results about primes requires a large amount of difficult techniques, from complex to Fourier analysis and so on.....but today we will show how imagination can lead us to a beautiful journey inside prime numbers!
I will review every step of his proof, which in my opinion is very creative and ingenious.
Do you know other marvellous elementary proofs? Let me know in the comments 😎
I apologize for the bad english but I have to admit that I'm italian, so maybe it is understandable 😜
❗❗❗ DISCLAIMER: at about 17:20 I use the bound 4^x and not 4^(x-1). This is in fact a weaker bound than the one given by the third key lemma, but I will use it anyway only because I want to speak about the Landau's trick, which works very well in the range of primes in [2,4001]. ❗❗❗
For other videos in English look at the following playlist:
ru-vid.com/group/PLuyidMWlEjrqmoFTxxig33J2IOeZItrzK
------------------------------------------------
00:00 Introduction
01:44 Proof of First Lemma
04:50 Proof of Second Lemma
09:18 Proof of Third Lemma
12:50 Final proof+Landau's trick
19:25 Conclusion
------------------------------------------------
#3Blue1Brown #SoMe1

Наука

Опубликовано:

 

22 авг 2021

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 46   
@zygoloid
@zygoloid 2 года назад
@14:34 We're assuming that p and 2p | (2n)! (and 3p does not divide (2n)!) implies that p²||(2n)!, but that's not true when p=2. Indeed, with p=2, p|2n C n when ⅔n
@MATHsegnale
@MATHsegnale 2 года назад
An amazing video for an amazing result! Best wishes for #some1 😉
@irrazionalex226
@irrazionalex226 2 года назад
Thank you guys! 😎
@leif1075
@leif1075 2 года назад
@@irrazionalex226 why do you say this is elementary?? IF someone didn't know this beforehand I don't see why would anyone would.think of this tonsolve this postulate..even a math whiz like Ramanujan or Russell or me..ornwhybwould anyone think of the postulate or these tools in the first place?
@thbeam2044
@thbeam2044 Год назад
Great video, was struggling to understand and this helped me, thanks!! (I think the 'i' you write at 8:01 is maybe meant to be a '2'? Just a small typo though :) I was a bit confused for a while.)
@liyi-hua2111
@liyi-hua2111 2 года назад
overall it’s excellent. though it would be a minor problem, but i never learned that product over an empty set is 1. (or said that if statement was false then there is no p inside and then erase the term) also where can i find some introductions of landaus trick
@light_rays
@light_rays 2 года назад
wow! that was so cool
@cocoviz6863
@cocoviz6863 Год назад
How did the application of the formula got that resul at 8:13?
@eukleidesk6759
@eukleidesk6759 2 года назад
Nice! Thank you.
@irrazionalex226
@irrazionalex226 2 года назад
Thank you!
@Preparazione2punto0
@Preparazione2punto0 2 года назад
Wow! Good luck ;)
@irrazionalex226
@irrazionalex226 2 года назад
Thank you :)
@DarinBrownSJDCMath
@DarinBrownSJDCMath Год назад
Thank you for giving a CORRECT proof!
@eggtimer2
@eggtimer2 2 года назад
One more question please: your first key observation: P > (2n)^.5 even if P > [2n chose n], is that still P|| [2n chose n] by some definition?
@irrazionalex226
@irrazionalex226 2 года назад
In this case if P>[2n chose n] then P automatically does not divide the binomial coefficient. Still in general there are some primes strictly larger than sqrt(2n) and less than [2n chose n].
@ScissorstheClown
@ScissorstheClown 2 года назад
An Italian wearing a colorful headband. A book on the shelf that says "Dio".
@theadoenixes3611
@theadoenixes3611 Год назад
Can anybody please tell me how P^(alpha) is at most 2n ? How that formula concludes this result?
@a0z9
@a0z9 2 года назад
Erdos consiguió demostrar este resultado importante de la teoría de números ,que abre la posibilidad de demostrar otras cosas aún más importantes en teoría de números.
@kobethebeefinmathworld953
@kobethebeefinmathworld953 2 года назад
Beautiful proof!
@irrazionalex226
@irrazionalex226 2 года назад
Thank you very much!
@leif1075
@leif1075 2 года назад
@@irrazionalex226 Omg this is ALL RIDICULOUS CRAZY TRICKS HOW I don't see why Anyone would think any of this AT ALL..can you please respond when you can??
@eggtimer2
@eggtimer2 2 года назад
I like it more and more! Is q=2m+1 prime or not prime. Couldn't get it from the audio ...
@irrazionalex226
@irrazionalex226 2 года назад
Yes sorry for the induction it is sufficient to ask for q=2m+1 being prime
@eggtimer2
@eggtimer2 2 года назад
@@irrazionalex226 no need to apologise! I have to thank you for your help and patience!
@malifraz8945
@malifraz8945 Год назад
molto bene
@georgelaing2578
@georgelaing2578 Год назад
Thank you for a clear and complete treatment of this very interesting theorem.
@TI5040
@TI5040 2 года назад
I believe it's the proof of Paul erdos( It's one of elementary proofs)
@irrazionalex226
@irrazionalex226 2 года назад
Yes It Is exactly his proof 😎
@Spacexioms
@Spacexioms 2 года назад
saved for later
@irrazionalex226
@irrazionalex226 2 года назад
Thank you Juan!
@mncubing8160
@mncubing8160 2 года назад
*Taking notes for use in math Olympiad
@eggtimer2
@eggtimer2 2 года назад
Hmmm, there are some inconsistencies. I like it overall..
@irrazionalex226
@irrazionalex226 2 года назад
Thank you! Yes maybe I've said something incorrect, if you want you can write here the inconsistencies and I'll be glad to explain them better
@eggtimer2
@eggtimer2 2 года назад
@@irrazionalex226 you did a great job. Wondering why you got the 2+ out in the first lemma Just to pitnit back in?
@irrazionalex226
@irrazionalex226 2 года назад
@@eggtimer2 yes in order to have in the final formula 2n and not (2n+1). In fact an easier computation show that (by a mean-inequality type of argument) (2n,n)>4^n/(2n+1)
@eggtimer2
@eggtimer2 2 года назад
@@irrazionalex226 Tha ks for your reply. But now I am confused: 2n was your starting point? Can you explain a bit more why that is. Appreciate that I take a lot of your time but really appreciate it.
@irrazionalex226
@irrazionalex226 2 года назад
@@eggtimer2 yes the main idea behind this proof is to reach a contradiction by bounding the binomial coefficient C(2n,n) in two ways, showing that for every big enough n these two bounds are incompatible (assuming that Bertrand's postulate is false). In this way the first lemma gives exactly the first bound for the binomial coefficient we are looking for.
@AdolphusOfBlood
@AdolphusOfBlood 2 года назад
are you not counting 1 as a natural number?
@irrazionalex226
@irrazionalex226 2 года назад
Sorry could you point the minute in which I've said that?
@deepp0
@deepp0 Год назад
the music is annoying thanks for the vid!
@leif1075
@leif1075 2 года назад
I don't see why on Earth ANYONE WOHLD EVER think of ANY of this if you were told this postulate and asked to.prove it..the first step especially comes out of nowhere..and I'm very good at math..
@ahoj7720
@ahoj7720 Год назад
As for many proofs this is presented in reverse. The starting point is the remark that binomial(2*n, n) is an integer, then exploring its prime factors. According to Erdös himself the key remark is that it cannot have factors between 2/3n and n. And I agree with you: Erdös was not anyone!
@DarinBrownSJDCMath
@DarinBrownSJDCMath Год назад
Erdös likely discovered this proof not by seeking to find a new proof of Bertrand's postulate, but just by playing around with the prime factorizations of C(2n, n) and trying to see what he could get out of it.
@DavideCerriGA
@DavideCerriGA 2 года назад
Neck beard and bandana. Are we related?
@amazinggadgets7554
@amazinggadgets7554 10 месяцев назад
Very bad quality of video