Question 313163
Check out the link below for the particulars.
You can do this all within an EXCEL spreadsheet. 
You're trying to fit {{{y=a0+a1*x+a2*x^2}}}to the given {{{N}}} sets of data ({{{xi}}},{{{yi}}}).
.
.
.
You need to create a matrix of values.
[X]={{{(matrix(3,3,a,b,c,b,c,d,c,d,e))}}}
where 
{{{a=N}}}
{{{b=sum(xi,1,N)}}}
{{{c=sum(xi^2,1,N)}}}
{{{d=sum(xi^3,1,N)}}}
{{{e=sum(xi^4,1,N)}}}
.
.
.
Then create another matrix of values,
[Y]={{{(matrix(3,1,p,q,r))}}}
where
{{{p=sum(yi,1,N)}}}
{{{q=sum(xi*yi,1,N)}}}
{{{r=sum(xi^2*yi,1,N)}}}
Then you have the coefficients a0,a1, and a2. 
[a]={{{(matrix(3,1,a0,a1,a2))}}}
.
.
.
The matrix equation you need to solve is,
[X][a]=[Y]
[a]=[X]inv[Y]
So find the inverse of [X] and matrix multiply by [Y]. 
That will provide you the coefficients of the least square fit quadratic, [a].
.
.
. 
http://mathworld.wolfram.com/LeastSquaresFittingPolynomial.html