Friday, 4 May 2018

combinatorics - Inductive proof for binom2nn=sumlimitsnk=0binomnk2



I want to prove the following identity using induction (not double counting method). Although it is a specific version of Vandermonde's identity and its inductive proof is presented here, but I need a direct inductive proof on this, not the general form.



(2nn)=nk=0(nk)2




I have tried to simplify (2n+2n+1) using Pascal's theorem, but did not get any result. Any help?


Answer



Suppose
nk=0(nk)2=(2nn)


then
n+1k=0(n+1k)2=n+1k=0[(nk)+(nk1)]2=n+1k=0[(nk)2+(nk1)2+2(nk)(nk1)]=(2nn)(1+1+2nn+1)=(2n+2n+1)

Explanation:
(2a): Pascal's Triangle identity
(2b): algebra
(2c): apply (1) and (3)
(2d): (2n+2n+1)=4n+2n+1(2nn)







Lemma:
nk=1(nk)(nk1)=nn+1nk=0(nk)2


Proof:



Since (nk1)=knk+1(nk), we have (nk)+(nk1)=n+1nk+1(nk). Therefore,
nk+1n+1[(nk)+(nk1)](nk1)=(nk)(nk1)



Since (nk)=nk+1k(nk1), we have (nk)+(nk1)=n+1k(nk1). Therefore,
kn+1[(nk1)+(nk)](nk)=(nk1)(nk)

Adding (3a) and (3b) and cancelling yields
nk+1n+1(nk1)2+kn+1(nk)2=(nk1)(nk)

Summing (3c) over k, and substituting kk+1 in the leftmost sum, gives
nn+1nk=0(nk)2=nk=1(nk1)(nk)

QED


No comments:

Post a Comment

real analysis - How to find limhrightarrow0fracsin(ha)h

How to find limh0sin(ha)h without lhopital rule? I know when I use lhopital I easy get $$ \lim_{h\rightarrow 0}...