Solving An INSANELY Hard Viral Math Problem
HTML-код
- Опубликовано: 1 авг 2024
- This seemingly simple viral problem is a lot harder than it looks--it is actually a problem from a university level mathematics textbook! In order to solve the problem, we take a journey through symmetry and group theory which leads to a simple formula for solving these kinds of equations.
Thanks to all patrons! Special thanks this month to: Kyle, Michael Anvari, Richard Ohnemus, Shrihari Puranik.
0:00 Problem
1:44 Numerical solution
2:37 Symmetric polynomials
4:15 Power sum polynomials
5:03 Elementary symmetric polynomials
5:46 Deriving Girard-Newton identities
8:14 Girard-Newton identities
9:28 Solution
References/alternate methods listed in my blog post
wp.me/p6aMk-8z8/#sources
Multiply by lines and introduction to group theory
• I discovered a new way...
Subscribe: ruclips.net/user/MindYour...
Send me suggestions by email (address in video). I consider all ideas though can't always reply!
Like many RUclipsrs I use popular software to prepare my videos. You can search for animation software tutorials on RUclips to learn how to make videos. Be prepared--animation is time consuming and software can be expensive!
Why are there comments before the video is published? Get early access and support the channel on Patreon
/ mindyourdecisions
If you buy from the links below I may receive a commission for sales. (As an Amazon Associate I earn from qualifying purchases.) This has no effect on the price for you.
Show your support! Get a mug, a t-shirt, and more at Teespring, the official site for Mind Your Decisions merchandise:
teespring.com/stores/mind-you...
My Books (worldwide links)
mindyourdecisions.com/blog/my...
My Books (US links)
Mind Your Decisions: Five Book Compilation
amzn.to/2pbJ4wR
A collection of 5 books:
"The Joy of Game Theory" rated 4.1/5 stars on 45 reviews
amzn.to/1uQvA20
"The Irrationality Illusion: How To Make Smart Decisions And Overcome Bias" rated 3.5/5 stars on 4 reviews
amzn.to/1o3FaAg
"40 Paradoxes in Logic, Probability, and Game Theory" rated 4.1/5 stars on 14 reviews
amzn.to/1LOCI4U
"The Best Mental Math Tricks" rated 4.7/5 stars on 8 reviews
amzn.to/18maAdo
"Multiply Numbers By Drawing Lines" rated 4.3/5 stars on 6 reviews
amzn.to/XRm7M4
Mind Your Puzzles: Collection Of Volumes 1 To 3
amzn.to/2mMdrJr
A collection of 3 books:
"Math Puzzles Volume 1" rated 4.4/5 stars on 13 reviews
amzn.to/1GhUUSH
"Math Puzzles Volume 2" rated 4.5/5 stars on 6 reviews
amzn.to/1NKbyCs
"Math Puzzles Volume 3" rated 4.1/5 stars on 7 reviews
amzn.to/1NKbGlp
Connect with me
My Blog: mindyourdecisions.com/blog/
Twitter: / preshtalwalkar
Newsletter (sent only for big news, like a new book release): eepurl.com/KvS0r
2017 Shorty Awards Nominee. Mind Your Decisions was nominated in the STEM category (Science, Technology, Engineering, and Math) along with eventual winner Bill Nye; finalists Adam Savage, Dr. Sandra Lee, Simone Giertz, Tim Peake, Unbox Therapy; and other nominees Elon Musk, Gizmoslip, Hope Jahren, Life Noggin, and Nerdwriter. 
Наука
![J.P., NLE Choppa - Bad Bitty (Remix) [Official Music Video]](http://i.ytimg.com/vi/rBDa6BCvw10/mqdefault.jpg)
Special thanks to everyone who watched the video before it went public! I shared this video early in the "community" tab (see ruclips.net/video/f3indrh6GEI/видео.html for the original) and you guys gave some great feedback!
The result was this new video with 2 very, very minor changes. (It did take me 5 hours to make these edits; that's how insanely hard it is to make these math animations!). But I felt it was worth making the edits to be as accurate as possible. The changes were:
a) At 5:12 I changed the subscripts to 1=j1
Hey do you happen to have a discord server to connect with your fans?
Thank you for uploading Tawalkar
Again a great video
Okay
@ken kennedy Yeah Ken You got Right One
Tom and Jerry
This feels less about being a genius and more about having studied maths.
I didn't solve by this approach, and I could solve it. I solved it by simple algebra.
He said that at the beginning. You can solve it using clever tricks and simple algebra, however he introduced us to a systematic approach, which will solve every problem of this type.
@@stewartzayat7526 Oh is it? My bad then. I guess I just rushed straight for the solution. Sorry!
@@divyanshu30gupta you have nothing to apologise for
Divyanshu Gupta can you give me a hint please?
From the thumbnail, I thought that the question was asking for the value of xⁿ+yⁿ+zⁿ ; so I was looking for a closed-form solution for that expression. (At one point I even hesitated and wondered if this was a trick question, but the video was listed as over 12 minutes long, which made me believe there must be some very advanced legitimate solution.) When I had spent much time trying to figure out how to turn the recursive formula that I found into a simple closed-form solution, and I finally watched the video, it turns out that it only asks for x⁵+y⁵+z⁵.
Anyway, the closed-form solution is
xⁿ + yⁿ + zⁿ = cⁿ + (a+ib)ⁿ + (a-ib)ⁿ ,
where a,b,c are real constants, and where c is the real root and {(a+ib), (a-ib)} are the two complex roots of the cubic equation
q³ = q² + (1/2)q + (1/6)
which is the characteristic equation of the recursive formula
p[n+3] = p[n+2] + (1/2)*p[n+1] + (1/6)*p[n]
Due to the recursive formula (with rational coefficients and rational start values), the value of xⁿ + yⁿ + zⁿ for integer n must be rational. Since the absolute values of (a+ib) and (a-ib) turn out to be less than 0.5 , the (a+ib)ⁿ and (a-ib)ⁿ terms vanish for increasing values of n . This and the fact that the value for integer n is rational, means that we may find an alternative exact closed-form solution for integer n that uses a rounding function instead of these complex roots (a+ib) and (a-ib). This exact closed-form solution (only for integer n) is:
xⁿ + yⁿ + zⁿ = F(n) = round( h(n)*c^n )/h(n) ,
where
c = [ 2 + ³√(44+6√26) + ³√(44-6√26) ]/6 = 1.43084957... is the real root of q³ = q² + (1/2)q + (1/6),
round(X) is the rounding function which rounds X to the nearest integer,
h(n) is a factor that depends on the value of m = (n mod 6) and k = floor(n/6) :
n = 0 + 6k ==> h = (1/6) * 72^k
n = 1 + 6k ==> h = (1/2) * 72^k
n = 2 + 6k ==> h = 1 * 72^k
n = 3 + 6k ==> h = 1 * 72^k
n = 4 + 6k ==> h = 6 * 72^k
n = 5 + 6k ==> h = 6 * 72^k
This formula works for integers n > 3 . So, some examples to see this formula in action:
For n = 5, m = (n mod 6) = 5, k = floor(5/6) = 0 , and thus
x⁵+y⁵+z⁵ = F(5) =
= round( h(5) * c⁵ ) / h(5)
= round( 6*72⁰ * (1.43084957...)⁵ ) / [6*72⁰]
= round( 6 * 5.99749489... ) / [6]
= round( 35.9849693... ) / 6
= 36/6
= 6
For n = 6, m = (n mod 6) = 0, k = floor(6/6) = 1 , and thus
x⁶+y⁶+z⁶ = F(6) =
= round( h(6) * c⁶ ) / h(6)
= round( (1/6)*72¹ * (1.43084957...)⁶ ) / [(1/6)*72¹]
= round( 12 * 8.58151299... )/ [12]
= round( 102.978156... ) / 12
= 103/12
= 8.5833333...
For n = 13, m = (n mod 6) = 1, k = floor(13/6) = 2 , and thus
x¹³+y¹³+z¹³ = F(13) =
= round( h(13) * c¹³ ) / h(13)
= round( (1/2)*72² * (1.43084957...)¹³ ) / [(1/2)*72²]
= round( 2592 * 105.371146... )/ [2592]
= round( 273122.01... ) / 2592
= 273122/2592
= 136561/1296
= 105.371142...
(Note: these answers are _exact_ , not approximations.)
Anyway, thanks for another challenging math puzzle!
[ EDIT: Edited to fix an error in the recursive formula (I had left out the factor p[n] in the last term). ]
this is nice Master
how did you derive those equations and why are they not approximations although you are using the rounding function?
@ImPresSiveXD ,
Which formula or expression do you want to see the derivation of? There are several formulas and expressions in my comment, and giving the derivation of each of them would be too much for just one reply (and would possibly also not be what you want).
To see why the final formula is exact, even though it uses a rounding function, I'll illustrate the principle behind it with a simpler case: Fibonacci numbers.
A reminder of what Fibonacci numbers are: you start with F[0] = 0 and F[1] = 1, and every number after that is the sum of its two preceding numbers: F[n+2] = F[n+1] + F[n]. So the sequence of Fibonacci numbers starts 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89,...
The n'th Fibonacci number F[n] can be determined by the formula
F[n] = round( (1/√5)*φⁿ )
where
φ = (1+√5)/2 = 1.61803399... is the so-called _golden ratio_ , and
round(X) is the rounding function that rounds X to the nearest integer.
This formula for F[n] is _exact,_ even though it uses a rounding function. The reason why it works, is this:
The exact closed-form formula of F[n] (without rounding function) is given by
F[n] = (1/√5)*φⁿ - (1/√5)*(-1/φ)ⁿ
for any integer n ≥ 0
(you can probably easily look this formula up on Wikipedia or something like that). Now consider the second term, which is (1/√5)*(-1/φ)ⁿ = (0.447213595...)*(-0.618033989...)ⁿ . For n=0 its value is (0.447213595...), and for increasing n this value is repeatedly multiplied by (-0.618033989...). So the absolute value of this term will always be less than 0.5 .
So that means that the first term, which is (1/√5)*φⁿ , is always less than a distance of 0.5 removed from F[n]. But we also know that F[n] must be an integer (because that's obviously the nature of Fibonacci numbers; if you start with two integers, then each subsequent "sum of two preceding numbers" will also be an integer). So what the addition of the second term (1/√5)*(-1/φ)ⁿ essentially does, is bring the value of the first term (1/√5)*φⁿ to the nearest integer.
In other words: we may replace the addition of the second term with the operation of rounding the first term to the nearest integer, since they are the same thing. And that's how we obtain the Fibonacci formula that uses the rounding function.
The same principle also applies in my formula for p[n] (with p[n] being short-hand for xⁿ + yⁿ + zⁿ) , except (since p[n] is rational instead of just an integer) we're now applying the rounding function to the numerator of p[n] (because we know that that numerator must be an integer).
So suppose that we'd have expressions g(n) and h(n) such that p[n] = g(n)/h(n) , and g(n) and h(n) are integers (and not necessarily reduced), then we can write
h(n)*p[n] = g(n)
... substitute p[n] = cⁿ + (a+ib)ⁿ + (a-ib)ⁿ ...
h(n)*cⁿ + h(n)*(a+ib)ⁿ + h(n)*(a-ib)ⁿ = g(n)
We defined g(n) to be integer. So if we can show that |h(n)*(a±ib)ⁿ| < 0.25 (which is indeed the case with the proposed h(n) in my comment), then the addition of the second and third term in the lefthandside is equivalent to rounding the first term to the nearest integer:
round( h(n)*cⁿ ) = g(n)
And then we'll have
p[n] = g(n)/h(n)
= round( h(n)*cⁿ ) / h(n)
I hope that helps.
@@yurenchu Thank you very much that helps. But how did you derive h(n) ? The 72 seems strange, I know it`s 2*(6)^2 but why would we choose this value?
@ImPresSiveXD , You're welcome. I determined h(n) as follows.
The recursive relation that can be derived for p[n] = xⁿ+yⁿ+zⁿ , is
p[n+3] = p[n+2] + (1/2)*p[n+1] + (1/6)*p[n] ....{eq.1}
Given start values p[1] = 1, p[2] = 2, p[3] = 3, we can find p[4], p[5] etc. recursively by using this equation. Since this recursive equation has rational coefficients and since the start values are integers, we know that each subsequent p[n] will be rational (i.e not irrational). In other words, each p[n] can be written in the form p[n] = g/h , where g and h are integers (and g and h are not necessarily co-prime). Furthermore, since the recursive relation divides only by 2 and/or 6, we know that we can define denominator h to be of the form (2^α)*(3^β) , where α and β are nonnegative integers.
We could define
α(n+3) = max{ α(n+2), 1+α(n+1), 1+α(n) }
β(n+3) = max{ β(n+2), β(n+1), 1+β(n) }
for integer n > 3 . This matches the recursive description of p[n+3] : from the three terms in the righthandside of eq.1 , the term whose denominator (when factorised into prime factors) contains the most instances of prime factor 2, determines the quantity of prime factor 2 in the denominator of the righthandside sum, and hence also in the denominator of p[n+3]. And likewise with prime factor 3.
Starting with α(1) = α(2) = α(3) = 0 and β(1) = β(2) = β(3) = 0 (because the three start values of p[n] are integers, and hence have h = 1), we can determine the following table
n : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 etc.
α(n) : 0 0 0 1 1 2 2 3 3 4 4 5 5 6 6 etc.
β(n) : 0 0 0 1 1 1 2 2 2 3 3 3 4 4 4 etc.
We see that after n = 4, α(n) is increased with 1 after every two steps, and β(n) is increased with 1 after every three steps. (Note to the reader: prove that this pattern continues indefinitely for increasing n.) Generally, for n > 3,
α(n) = floor((n-2)/2)
β(n) = floor((n-1)/3)
From these expressions, we determine h(n) = 2^α(n) * 3^β(n) :
n = 0 + 6k ==> h = 2^(3k-1) * 3^(2k-1) = (8^k)/2 * (9^k)/3 = (1/6) * 72^k
n = 1 + 6k ==> h = 2^(3k-1) * 3^(2k-0) = (8^k)/2 * (9^k)/1 = (1/2) * 72^k
n = 2 + 6k ==> h = 2^(3k+0) * 3^(2k+0) = 1*(8^k) * 1*(9^k) = 72^k
n = 3 + 6k ==> h = 2^(3k+0) * 3^(2k+0) = 1*(8^k) * 1*(9^k) = 72^k
n = 4 + 6k ==> h = 2^(3k+1) * 3^(2k+1) = 2*(8^k) * 3*(9^k) = 6 * 72^k
n = 5 + 6k ==> h = 2^(3k+1) * 3^(2k+1) = 2*(8^k) * 3*(9^k) = 6 * 72^k
So there we have it.
I have two degrees in EE. I once helped create something called a gamma ray resonance spectrometer. I get paid to optimize GPU code. I've taught several graduate & undergrad classes in math-focused areas. I still can't do a lot of the problems that come up on your channel, and each explanation is *very* useful. Anyone that complains... well, let them go be themselves, and the rest of us can find education and enjoyment in symmetry and solutions. (Long time fan, first time caller. Keep up the great work.)
Rather than solving anything, this video just felt like "if you know these very specific identities, which we will neither prove nor explain, we can derive the answer directly out our own asses."
The identities were derived from very basic assumptions.
@@natewright1197 aka our ass
@@Gautam-tk8tf very basic ass umptions
@@chaotickreg7024 missed that lmao
1% seems optimistic. Maybe 0.01%
This is a beautiful solution, but by simply multiplying the first equation with the others (and itself) and keeping track of the terms, one can find the correct values. It is very prone to error. I would like to point out that whenever you are faced with an equation, you need to justify your roots, as to weed out false roots. In this case, the roots are non-real, so a solution simply doesn't exist over R.
Even 0.01% seems a little optimistic within a reasonable timeframe and without doing research
1% of mathematicians seems more likely to me.
@@mr.moodle8836 No, it is not that difficult. By applying basic maths and multiplying the given equations, the answer can be found. The only thing is, you have to be careful and patient because the solution is quite lengthy.
@@apoorvvarshney3276 The problem is that you have to develop identities to solve the question. Using the elementary symmetric polynomials almost feels like a random step, there isn't much to prompt most people into doing it in the first place (unless they already know of the Newton-Girard formulas, but the assumption is that you aren't aware of them and have to derive the formula as shown in the video to apply it, which I probably should've been clearer about, but I think that's a reasonable assumption for a problem-solving question). I'm sure more people could do the question, given they were given a little nudge, but otherwise, probably not. If you consider the entire developing world, and acknowledge that most people with a math degree probably hold a degree in some form of applied mathematics (hence probably wouldn't specialise in solving a problem from "abstract algebra"), it seems reasonable to say that less than 0.01% have experience working with this kind of maths. Now, I imagine a far smaller proportion of non-pure math majors would be able to solve this than the proportion of math majors who can't (Ds get degrees yo), so under 0.01%, including the ENTIRE world population, not just developed countries, is a perfectly reasonable estimate.
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
Very nicely and well explained!
Now I know that
e2 = ab+bc+ca
e3 = abc.
Thank you!
That's basically what he showed in the video though. Recall what exactly e_n and p_n mean.
It's very good bro 💪
But it's not *quite simple* 😂
Me after watching this video:
I am starting to appreciate my intelligence in other things that aren't math.
This is surely one of the more difficult problems on your channel. I'll have to watch it at least one or two times more to really understand it.
I just have fibbed this like "it seems 5, so it's 6"
I am in %1.
No you did not solve it
@@ZoneEEEEEEEEEEEE humour.
@@AmitGarnaik unfunny.
@@ZoneEEEEEEEEEEEE opinions vary.
@@AmitGarnaik i know, just stating mine
Presh, thanks for doing this.l usually binge watch your videos whenever I'm free. They are absolutely amazing. We might need a Netflix series soon!
0:29 after watching this again and again , I realised, i was too in the 99%🙃
I don't think even 1% could solve this one. With two maths 'A'-levels and an engineering degree (a long time ago, admittedly) I found this video way over my head compared to most of your videos.
Maybe because group theory is hardly ever covered in engineering math courses...nevertheless, imho one great thing about internet is this chance we have to learn new topics (almost for free) with great math channels from all around the world like this one, bprp, mathologer, 3b1b, numberphile, etc.
Not 1% generally but 1% of those who try which is a highly selective process
I'm currently studying an engineering major and I'm at algebra 3 now but I've never heard of the method. I think this is the theoretic approach of math which hasn't much to do with application which is we usually learn.
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
@@sajitsama7764 What you wrote,my friend,is essentially the internal working we did when we used the Girard-Newton identities and I admit,I did it the same way when I solved it. :)
Also from India ❤
8:37 kek
If u press on 8:37 and keep pressing it it sounds like he is saying nice in a weird way
enis enis enis enis enis enis enis
@@farbodzandinia niece
Ya jfjmfnfjfjdkdkd
This was a great topic and very clearly presented. Kudos. By the way, I’ve been a fan and supporter for a long time. I’ve learned so much from your videos over the years. I hope you never stop making them.
Brilliant, what a great piece of teaching. I knew enough about symmetric polynomials to blunder out a semi-numerical solution, involving finding the roots of a quadratic and then a cubic, and got the answer a lot closer than wolfram. However, Newton-Girard goes into my mental card-file
All I understand is-POLYNOMIALS
@@nabanildas5819 he's amazing..but I have no idea of permutations and didn't find it on his channel
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
Sajit Sama beautiful approach
Sajit Sama How are you sure there is no other value rather than 1 in front of the factors? I mean, it could be p(x)= A(x-a)*(x-b)*(x-c) with A≠1
[sorry if it is obvious, i have just learned polynomial concepts]
@Renno Vilela , the value of your parameter A doesn't affect the roots of p(x); so eventually, this value will cancel in the analysis anyway.
The question is asking for the value of x⁵+y⁵+z⁵ , not for the value of A(x⁵+y⁵+z⁵) .
I MIND MY DECISION now.
Hey do u like me
Tom and Jerry
@@Unpopular_Trader Im trying to like your comment but the section of the screen where the like button is does not work, sorry.
@@pepehimovic3135 No probs
@Adam Romanov Hey be Happy...
Don't be angry....
I am one who makes everyone happy....
@Adam Romanov And Enjoy Every Moment Of Life It Is Too Short To Enjoy...
This channel is really awesome.
I am a high school student and love to watch these videos in my free time.
It's really fun.
You completely explained the Girard-Newton Identity for solving that problem! This is what makes a channel exceptional. I am a school going student and could precisely understand everything you explained. Lots of love from Bangladesh! Keep going! ❤️💯🇧🇩
This is really superb MYD! Thanks for introducing advanced less-often known identities and concepts to masses. Cheers!
This is the first time I'm so excited about your video! I will memorize these facts!
I was well on my way to not understanding another amazing video explaining things I gave up on trying to understand way back in high school in Mr Backer’s algebra class, when I spent a week’s worth of after school tutoring in quadratic equations because I didn’t want to give up on it without a fight. At the end of that week Mr Baker politely informed me that I’d never understand them. The reason: there’s only so much stuff you can memorize without understanding at least some of it.
What sealed the deal in this video was the commercial in the middle of it. It wiped away any entertainment value for me, and I only finished watching it out of the desire to be polite.
I never took any college algebra courses, only calculus, so I was so lost this entire video until the last summation with the arithmatic. I have no clue how you found the formulas to get the answer, but it was fun to watch you work. It was a lot of e, i p, k, and I am accustomed to XY and Z 😂
Sum of 6th power is (103/12)
got that as well, then 221/18, 1265/72, 905/36,15539/432 etc
i got 17/4
@@nikolasscholz7983 Correct!
Matlab code:
syms x y z
solution = solve(x+y+z-1, x^2+y^2+z^2-2, x^3+y^3+z^3-3);
for i = [5:20]
s = simplify(solution.x.^i + solution.y.^i + solution.z.^i);
disp([i s(1)]);
end
Solution:
[ 5, 6]
[ 6, 103/12]
[ 7, 221/18]
[ 8, 1265/72]
[ 9, 905/36]
[ 10, 15539/432]
[ 11, 11117/216]
[ 12, 21209/288]
[ 13, 136561/1296]
[ 14, 260531/1728]
[ 15, 559171/2592]
[ 16, 9601075/31104]
[ 17, 6868847/15552]
[ 18, 39313147/62208]
[ 19, 84376799/93312]
[ 20, 482922025/373248]
I also got 103/12
@@nikolasscholz7983 I agree
I remember the n=5 problem. My approach was to consider different ways to produce polynomials with 5 as the highest order (i.e. (x^4+y^4+z^4)(x+y+z), (x^2+y^2+z^2)(x^3+y^3+z^3) and so on), and use those results to get rid of certain types of terms from the expansion of (x+y+z)^5 (i.e. xy^2z^2, x^2yz^2,x^2y^2z) one by one, (after doing the same for n=4 to get the value of x^4+y^4+z^4). I wonder if this is actually the same idea, just with more sensible notation (my calculations got very messy).
that’s exactly how I approached it too. You can group together all terms with equal powers, like x^2y^2z plus the other two permutations to keep it managaeble. You’ll get a 5x5 dim matrix equation which is straightforward to solve by hand by suitably adding rows. This only requires basic math skills; not the knowledge of particular theorems. Having said that; it was great to learn about the properties of symmetric polynomials, afterwards!
I have just studied symmetric group at college and this is truly a clever and marvelous approach.
It's great to see someone who respects history lately. I take your word for it on the math stuff.
I agree
Blackpenredpen: I was 4 universes ahead of you!
I was searching for this comment lmao😂
quadruple universes.
..... *PERPENDICULAR* Universes.
Can you give me the link of tht video
This is what I was expecting: ruclips.net/video/bshqdj19YRA/видео.html an algebraic solution
And here was me, thinking I could get away with using substitution for the systems of equations...
I love this...
I actually learnt the proof sometime ago,but today,it's just straight forward .
I can now compute x⁶+y⁶+z⁶+... From a given polynomial if given (x,y,z...) are roots.
Thanks for the knowledge 🙇🏿♂️❤️
I watched till 6:05 , then I started minding my decision
that's the exact point when i paused the video and just gave up lmao
Me too
Thanks for the time stamp I will be now on my way
2:19 Is like Moe Greene's "I talked to Barzini" in Godfather. Voice randomly changes.
Phresh: "only 1% can solve this"
Blackpenredpen: "am I a joke to you"
1% of total population in this world
Who is black pen red pen
1%, not 1 person.
@Zero Speed
I also come in those 1%
Proof:-
ruclips.net/video/5RAJzvtlU6Y/видео.html
@@chauhan.739 wow I am so genius mom
I have to say, my first time watching this I thought it was another problem where you wither know the formula or you don't, which disappointed me. However, after watching again and really going through the derivation, it really wasn't that unreasonable to expect someone to arrive at this method. Great general solution.
This was absolutely delightful. I encountered another similar problem in a book full of mathematical puzzles, but they are not the p polynomials but sums of e polynomials. I took a crack at it and eventually gave up. Little did I know there's so much stuff behind it. Thanks and now I can sleep in peace.
When I read other channels' title as this I'm like " I come in that 1%"
When I read this channels title as this I'm like " who else is there in that 1% except for u"
I did 2 years of math after highschool and couldn't solve it. So... maybe people who did 3 years of math after highschool ? which is definitely less than 1% of the population.
... U can literally learn this from aops
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
@@sajitsama7764 wow so actually u come in that 1%😏😁
@@sajitsama7764 I've already seen your comment a couple times replaying other's comments. This it's really an easier to understand and more elegant approach to the problem. Please share it again as a normal comment so that others can find it easier than posting it in other comments.
I don't know if less or more than 1% can solve the problem, but this solution is worth sharing. I also thought about the first step of your solution, but I didn't try it, so I don't know if I would be able to solve it or not.
Lesser degrees equations can be solved without this method, I have solved this problem by some methamatical operations(sqauring both side, substitution ect)
Great review, I love the Group theory aspect!
Interesting generalization. I discovered a similar pattern in a different polynomial expression.
For this specific problem, by nature of symmetry, I got 3, third degree polynomial = 0.
I then multiplied each polynomial of x, y and z by one power at a time. Since RHS=0, LHS becomes a recursive expression of known expression. Rest is simplification.
Now answering the last part of your question, proving it rational: As long as we are adding/subtracting rational terms of power coefficients, intuitively it is a rational number. However, proving that irrational or complex may be necessary condition. Will try. 😊
Please make some videos on solving IMO problems!
Spoilers!!!!
*THE TITLE ISN'T CLICLBAIT*
@ehud kotegaro Maybe you're a genius
it is because far more than 1% of people did algebra at this level
@@martiniquevodka5574 really no not even close to 1‰
cliclbait..?
Hard
IMPORTANT QUESTION. Dear Pradesh, the solution is very clever. However there is a point which is not obvious and without explanation. Namely, you use en=0 for n>3, which applies to our 3 variables case, but the n-th relation between p and e coefficients (which you use to find p4, p5) is derived for an n-grade polynomial (resp. 4, 5), where there are n solutions and en (which is a function of these solutions) is obviously not 0 (it is actually x1x2x3 ...xn). I understand that it works, but I think that it deserves a complete explanation. Thanks. Samuele
This video was highly insightful... Loved it
My way of solving:
As X,y,z can be replaced and we know sum of X,y,z and xy,yz,zx and we also know xyz. Then make a cubic in t with X,y,z as root and multiply both sides by t know we get summation In t^4. So we can find x^4+y^4+z^4 by replacing t by X,y,z respectively and adding.
Agreed that it is a lot harder. I can visualize the 1st equation (plane) intersecting with the 2nd equation (spherical shell). Their intersection is a circle. Harder to visualize is the cubic in 3D (it's looks wavy like the bottom of an egg carton). I enlisted the help of mathematica to do a contour plot of all 3 graphs superimposed. If you rotate the graph, you can see that there is no intersection of the 3 equations in 3D space. Thus, the solution would involve imaginary numbers. Once you get to the complexity where you need mathematica to solve for x,y,z, you've taken away the relevance to 90% of the audience. Now that's just my opinion. The analysis to get the exact answer is wonderful, but not elegant. By elegant, I mean within reach and can be appreciated at the college level. It doesn't do it for me.
Great problem, and an even greater solution. Thanks, PT :)
without knowing these symmetric polynomials...
if you square the first eq and use the value of the second, you get:
(x+y+z)^2 = 1 = (x^2+y^2+z^2)+2(xy+yz+zx) = 2+2(xy+yz+zx)
so xy+yz+zx = -1/2
now cube the first eq, rearrange the terms to get expressions you know, and you get:
(x+y+z)^3 = 1 = (x+y)^3+3z(x+y)^2+3(x+y)z^2+z^3 = x^3+3yx^2+3xy^2+y^3+3zx^2+6xyz+3zy^2+3xz^2+3yz^2+z^3
(x^3+y^3+z^3)+3(x^2(y+z)+y^2(x+z)+z^2(x+y))+6xyz = 1
the first term is 3; y+z=1-x
3+3(x^2(1-x)+y^2(1-y)+z^2(1-z))+6xyz = 3+3(x^2+y^2+z^2)-3(x^3+y^3+z^3)+6xyz = 3+6-9+6xyz = 1
so xyz = 1/6
now let's consider p(n-1)
p(n-1) = p(n-1)*1 = (x^(n-1)+y^(n-1)+z^(n-1))(x+y+z) = x^n+x^(n-1)(y+z)+symmetrical terms = x^n+x^(n-2)(xy+xz)+symmetrical terms
[using xy+yz+zx = -1/2]
p(n-1)= x^n+x^(n-2)(-1/2-yz)+symmetrical terms = p(n)-p(n-2)/2-x^(n-3)*xyz+y^(n-3)*xyz+z^(n-3)*xyz = p(n)-p(n-2)/2-p(n-3)/6
[using xyz = 1/6]
so we get
p(n) = p(n-1)+p(n-2)/2+p(n-3)/6
we already know the first 5 p's: 1, 2, 3, 25/6, 6
p(6) = p(5)+p(4)/2+p(3)/6 = 6+(25/6)/2+3/6 = 103/12
p(7) = 221/18
p(8) = 1265/72
p(9) = 905/36
p(10) = 15539/432...
This is briliant.
I solved it with the elementary approach!! It felt very nice while reducing the last factor to 6
Nisarg Bhavsar Can you show your solution?
Nisarg Bhavsar It's very interesting.
I also solved this by elementary algebra:
1. By squaring the first equation you can find the sum x*y+x*z+y*z
2. By multiplying the first equation with the second you can find the sum x^2*y + x^2*z + y^2*x + y^2*z + z^2*x + z^2*y
3. By cubing the first equation you can find the value of x*y*z because the sum from step 2 is already known
4. Then you can get a system of linear equations for the sums z^5 + y^5 + z^5, x^4*y + x^4*z + y^4*x + y^4*z + z^4*x + z^4*y and x^3*y^2 + x^3*z^2 + y^3*x^2 + y^3*z^2 + z^3*x^2 + z^3*y^2. Notice, that sums like x^2*y^2*z + x^2*z^2*y + y^2*z^2*x can be factored like x*y*z*(x*y+x*z+y*z) so are already known. You can get three equations for the system by:
a) multiplying the second equation with the third equation
b) multiplying the first equation squared with the third equation
c) multiplying the first equation with the second equation squared
Solve that system and you will find z^5 + y^5 + z^5.
One of the hardest problems on this channel… even the solution confuses me 😅
i didnt undestand, but i solved it the "simple algebra" way. didnt understand this method, but only watched video once. watchign it again to understand
CB May be your best video it introduced a whole new concept to me I especially appreciated the elegance of this approach having slugged to the solution via high school algebra ala black pen
I did it with two variables changes, solving complex sinus and logs, using graphics to understand what i was doing and fighting with huge formulas with square roots inside of cubic roots..., and now i've watched your video and...... :)
Purely based on the mathematical content, I feel like this was one of your best videos to date.
It took me a few different approaches until I solved this myself using elementary algebraic manipulations. I had to solve a linear system of three equations in the end.
When I solved it, I realized this was one of the more difficult problems and I was curious what your approach would be. I am glad to say that your method of solving is much more interesting than mine. Very interesting method of solving!
Would you mind sharing with us your method?
I thought this is gonna be x^n+y^n+z^n=n so I used mathematical induction method.
Then failed beautifully.
Same
Thank you, I learned something new today.
The first puzzle of all your videos that I actually got right !
I remember I was able to solve a problem exactly like this while I was in high school. Expanding (x + y + z)^3 and doing some algebra there. Keep going to (x + y + z)^4 and then (x + y + z)^5. Nothing like the solution you presented here
That's not how it works!?!
@@pyro6631 there is a way to do it, i just commented a way to do it w/o any knowledge of complex number, i can repost it here if you want
@@starrmayhem repost
@@jaibhagwanyadav2387 just to warn you it would seem like magic if you don't solve it on your own
(x^2+y^2+z^2)(x+y+z) = (x^3+y^3+z^3) + (x^2*y+x^2*z)+(y^2*z+y^2*x)+(z^2*x+z^2*y)
(x+y+z)^3 = (x^3+y^3+z^3) + 3[(x^2*y+x^2*z)+(y^2*z+y^2*x)+(z^2*x+z^2*y)] + 6xyz
as you can see, there are terms that can be cancelled when 3*(x^2+y^2+z^2)(x+y+z) - (x+y+z)^3
which gives xyz = 1/6
let s(n) = (x^n+y^n+z^n)
s(n)*s(1) = s(n+1) + [(x^n*y+x^n*z)+(y^n*z+y^n*x)+(z^n*x+z^n*y)]
I hve also solved it by another method
This is harder than learning music notes lol, i got lost when my eye blinks hahaha
Thank you for the video! You friends are all super awesome! Oh, moments with this video are sad.
Best video so far on Newton Girard formulas
Everybody That solved it:
Presh: *they are the brightest group of people i have ever seen*
Uri Sadan-Yarchi חמר
@@shaunshuster7234 Thanks.
Nice! I actually didn’t know how to solve it, but my focus was pde, Fourier analysis, but I will try to solve it via linear algebra.
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
Sajit Sama thank you! That’s an elegant algebraic approach- love it ❤️
For middle or high school students who participate in math competitions, it is useful to learn how to solve this problem without remembering complex theories. To this end, one should start with
(x+y+z)[x^(n-1) + ...] = [x^(n) + ...] + [x^(n-2) + ...] (xy + xz + yz) - [x^(n-3) + ...](xyz). Using the given conditions, one can solve a=xy+xz+yz= -0.5 and b=xyz=(1/6). Then, one has
[x^(n) + ...] = [x^(n-1)+...] + [x^(n-2)+...]/2 + [x^(n-3)+...]/6.
As an interesting exercise, one may continue to find the equation to solve [x^(-1)+...], [x^(-2)+...], [x^(-3)+...], ...
I did this by assuming this might be an identity and I saw the symmetry. So I knew there are 6 solutions (may be not all unique). Since the problem did not ask for the roots ie x,y,z I considered to hold out and assume x=0 reducing all equations to two variables.
Then I multiplied one and two to get the value of y+z and then further multiplied two and three to get the required answer as 6.
In hindsight I was actually deriving values for the e's as mentioned in your video. It was great learning some new identities.
Or you can find a 3rd order difference equation for which x^n+y^n+z^n is a solution, and use the given values 1,2 and 3 as initial conditions.
Edit: I now see that this is essentially what you've done (but using a more algebraic approach)!
Spolier:
I believe the correct difference equation is I_n=I_{n-1}+I_{n-2}/2 + I_{n-3}/6. where |_n=x^n+y^n+z^n. The proof is by induction!
So I_4 = 3+2/2 +1/6 =25/6
I_5= 25/6 + 3/2 + 2/6 =36/6=6
I_6=6+25/12+3/6=103/12
I_7=103/12+6/2+25/36=442/36=221/18
etc...
Hi Benom8; very good!
Here is my post [for all]:
This problem is not that hard as it seems, and can be solved in a basic way! You just need to know a few formulas including those of Newton's binomial.
It leads to a 3rd degree equation in z (for example), which one can avoid solving [by Cardan's method, or by mine (visible on my draftWiki page)]; ...
... because the question just asks to determine the value of the expression E = x⁵ + y⁵ + z⁵ → which can be written as a function of the 3rd degree in z, and therefore be linked to the equation in z.
There are a few calculations, but nothing insurmountable.
Have a nice day, everyone :D
♥️♥️♥️♥️♥️
For those who want a longer yet a more general, 8th or 9th grade math, here's my method.
First, multiply (x+y+z)(x+y+z) and get the value of xy+xz+yz,
next, calculate (x2+y2+z2)(x+y+z) to get the value of xyz.
Then to calculate other value you just need to multiply the equation one order below with (x+y+z), Ex. to calculate (x4+y4+z4) you will multiply (x3+y3+z3)(x+y+z), (x5+y5+z5) needs (x4+y4+z4)(x+y+z), and so on.
MYD method are the simplification of these patterns. However, this way you only need one simple method, that is multiplying them, instead of remembering one more formula.
(the reason you need those 2 values is because all of the multiplication here could be simplified into xyz and xy+xz+yz)
Ex. For those who already calculate you'll know xyz=1/6, xy+xz+yz=-1/2.
Then x4+y4+z4= 25/6.
So x5+y5+z5 can be calculated like:
(x4+y4+z4)(x+y+z)=25/6*1
(x5+y5+z5)+xy4+yx4+xz4+zx4+yz4+zy4=25/6
f(5)+xy(x3+y3)+xz(x3+z3)+yz(y3+z3)=25/6
f(5)+xy(3-z3)....=25/6
f(5)+3(xy+xz+yz)-xyz(x2+y2+z2)=25/6
f(5)+3(-1/2)-(1/6)(2)=25/6
f(5)-11/6=25/6
f(5)=x5+y5+z5=36/6=6, and there u go.
Additional reply :
The answer might be able to be simplified as f(5)+f(3)*(xy+xz+yz)-f(2)*xyz=f(4)*f(1)
or f(n)=f(n-1)f(1)+f(n-3)*xyz-f(n-2)*(xy+xz+yz)
You can simplify the x,y,z into f(x) again but i'm too tired and not worth it, as at this point MYD's way is simpler ( maybe, too tired)
Amazing result.
The main utility of the Girard-Newton identities is that the roots do not need to be calculated. The roots only approximate the answer which is rational. This is covered in the chapter on Galois theory.
After the video ends I'm still sitting there eating my pie as idk what is going on in my monitor
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
@@sajitsama7764 are you speaking english? I havent learned most of this things (Im not the smartest person ever tho). Where did you learn this?
@@alex2005z I've learnt this in high school!
@@sajitsama7764 how old were you when you learned this? Just to have an idea my countrys school sistems is different then americas
@@alex2005z 15 years old probably!
again..your solution is great sir!
Thank you for teaching us about the Girard-Newton polynomial equations. I'd learnt in school that in a monic nth-degree polynomial, e_k, the sum of the products of the roots taken k at a time, is (-1)^k a_{n-k}, but not about G-N. I feared that the solution for x^5+y^5+z^5 would be some arbitrary degree-5 combo of e_0,...,e_5 which nobody could be fairly expected to find under exam conditions, but you showed us a systematic way. Thank you.
Explanation starts at 4:20.
6:21 Nobody going to note Presh saying "negative" for "minus"? Sounds like military or air-pilot jargon.
10:05 Respecting mathematics more than engineering, I'd prefer "-1/2" to "-0.5".
Me: Done advanced engineering maths at uni (thinks: "ah maybe i'll be able to follow the solution")
Next minute: (?).(?)
It is a previous years JEE Advanced question.
Was that a school for the gifted or a school in a city?
Yeah mate, we did
IIT JEE prep school
@@huytrandang277 no it's a famous and very good school in India
I too study there and i am 8th..😀
Thank you very much! You help me to solve lots of time!
Solved it. The idea is to express higher order polynomials in terms of lower order polynomials and cross-terms. Then, you use the given numerical values to solve for the cross terms. After seeing the video, I realized that what I did is equivalent to Girard-Newton.
Vídeo: 99% can't do it !!!
All youtube comments: i found the answer when i was sleeping and my dog took about 1 h to do it
This is the mathematics culture in a nutshell.
Excelente video, en mi intento de resolver ese problema pensé que salía 5 :'/
Ya luego aplique la extensión de Horner con las potencias de las raíces y me salió también
Ahora se un poquito más, Muchas Gracias ^-^ y excelente video
I did x3(X2) +y3(Y2)+z3(Z2) and got right answer. Because when basis are same powers are added.
A similar problem appeared on the STEP 3 math exam for the UK but for general n a few years back. A very fun problem.
A much more hardcore method, you can solve the system of 3 equations for unknown x,y,z using Vieta's theorem
That's what I was thinking, though I don't know what Viete's Theorem is. It doesn't seem necessary to do sums like this
@@umbragon2814 Vieta's theorem links a system of sums and products of N numbers to the coefficients of an N- degree polynomial (in this case a polynomial of deg 3)
@Adam Romanov Yes, it does. all you have to do is some substitutions to end up with x+y+z=1, xy+xz+yz=1/2, and xyz=1/6 which leads you (using Vieta's formula) to the equation X^3+X^2-1/2*X+1/6=0. All what's left is solving the 3rd deg eq with Cardaro method. Besides, I precised that this is "hardcore method", it is not optimal but, for instance, has other uses like proving that the system S(x,y,z) has no solutions in R^3
@@khaledshell507 6x^3 - 6x^2 -3x -1 = 0
There is no need to actually solve this equation.
It gives you a linear recursive formula for all higher powers of x:
x^n=(6x^(n-1)+3x^(n-2)+x^(n-3))/6,
and the same for y and z.
Now simply add the three recursion formulas and you get
p_n=(6p_(n-1)+3p_(n-2)+p_(n-3))/6,
where p_n is x^n+y^n+z^n.
Plug in the initial values 1,2,3 and you're done.
We can use reduction formula of these equations
How did wolfram alpha find complex values
Numerically solver.
Thank you bro... Best trick for polynomial 👍
I am a student of class 8th and when I saw this problem I solved it in an another way through only algebra and then using cubic formula to find the value of one variable and using it's value to find the other variables.
thank you for letting me know that I know nothing about Mathematics when I am 24
Never too late.
Nevah to latee
too little too late
Same here once i understand it i got a good grip but not paying much attention on high school, i had a ziroo idea on the video just watching it because it's interesting 😂
I saw this on BLACKPENREDPEN
Edit: what i thought : 1+2+3=6 :)
AaronThe Proj me too
Ahhhh yessss
Link, please
@@schwanky7 ruclips.net/video/1TBVeuOcY1w/видео.html
TB Cubing Thank you! I was having trouble finding it.
excellent solution
Great explanation
I solved it in different way and get the answer, I can't believe it was true))
I wish I was like u :'(
it's not easy for me to understand in such short time😂
The best thing about the video is that it is 12: 34
As 1,2,3,4...
This was actually one of my homework problems (exact same one) and you helped me understand it. I’m not even a Math major, why was this assigned to me :’)
I'm glad I didn't waste time trying to figure it out myself. I could have spent all day solving this and achieve nothing.
How about I calculate both a⁴+b⁴+c⁴ and a^5+b^5+c^5 by short method using only algebra.
Let if a,b and c are roots of a cubic polynomial f(x).
f(x) = (x-a)(x-b)(x-c) = x³ - (a+b+c)x² + (ab+bc+ca)x - abc
Now from the given equations.
a+b+c = 1
Let squaring both the sides
(a+b+c)²= 1²
Let's expand the left side
a²+b²+c²+2(ab+bc+ca) = 1
Now , we know that a²+b²+c²= 2 then
2(ab+bc+ca) = 1-2 = -1
Hence, ab+bc+ca = -1/2
And we know that
a³+b³+c³ - 3abc = (a+b+c)(a²+b²+c²-ab-bc-ca)
Putting all the values we have in the above equation we get,
3 - 3abc = (1) (2-(-1/2)
3 - 3abc = 5/2
3abc = 1/2
abc = 1/6
Hence our polynomial after substituting all these values will be
f(x) = x³ - x² - (1/2)x - 1/6
Now 'a' is the root of f(x),
Mathematically
a³ - a² - (1/2)a - 1/6 = 0
By multiplying with 'a' we get a⁴ ,
a⁴ = a³ + (1/2)a² + (1/6)a
We will get similar equations for b and c.
After adding them,
a⁴+b⁴+c⁴ = (a³+b³+c³) + (1/2)(a²+b²+c²) + (1/6)(a+b+c)
= 3 + 2/2 + 1/6 = 25/6
Similarly, if we multiply with 'a²' we get equation for 'a^5' i.e.
a^5 = a⁴ + (1/2)a³ + (1/6)a²
Adding again,
a^5 + b^5 + c^5 = 25/6 + 3/2 + 2/6 = 6
That's it! Quite simple
Love from India❤️
@@sajitsama7764 Awesome solution! I solved it too but with a waaayyy longer way. More like a dumber version of your solution lol. Yours is really good though great job! Makes me question my mathematical skills
@@denizkaragullu6239 thank you!
@@sajitsama7764 I solved in the same way.
5:10 "give you some intuition" ... exactly the opposite of intuition
Very complex to understand. Almost a classroom lecture. Ok with some tricky solution explained for so many earlier problems
Well explained
The recursion is: Sum of the nth powers = the sum of the (n-1)st powers + 1/2*(sum of the (n-2)nd powers) + 1/6(sum of the (n-3)rd powers), so they'll all be rational and the sum of the 6th powers is 6 + 1/2 * 25/6 + (1/6)*3 = 103/12. Sheesh, this is so easy. I hear they give this problem to 3rd graders in Singapore.
When I was at exactly in the middle of the video, me: let me see the comments once, who might have understood this??😞
Elementary approach using first principles as follows useful for those who have not exposed to higher Algebra.
by squaring (x + y + z) and applying the known result x^2 + y^2 + z^2 = 2; we get xy + yz + zx = -1/2
Then multiplying (x + y + z)(x^2 + y^2 + z^2) and applying above results we get xyz = 1/6
By squaring (xy + yz + zx) we get (x^2)(y^2) + (y^2)(z^2) + (z^2)(x^2) = -1/12
By squaring x^2 + y^2 + z^2 we get x^4 + y^4 + z^4 = 25/6
Then multiplying (x^3 + y^3 + z^3)(x^2 + y^2 + z^2) and applying above results we get x^5 + y^5 + z^5 = 6
Squaring (x^3 + y^3 + z^3) and applying above results we get x^6 + y^6 + z^6 = 103/12
does (x^3 + y^3 + z^3)^2(x + y + z) solve for x^7 + y^7 + z^7
you seem to be skipping the "how to apply the results part"
This channel made maths lovely