Here it is matlab-script I build to represent data into linear equation.

function [A] = linierfit(xi,yi)
% regresi linear
% y = a0 + a1x

nxi = numel(xi);
nyi = numel(yi);
if nxi ~= nyi
    disp('ukuran data tidak sama');
    return;
end
n = nxi;
sxi = sum(xi);
xi2 = xi.^2;
sxi2 = sum(xi2);

%matrix X
sigmax = [n sxi;sxi sxi2];

%matrix Y
syi = sum(yi);
xiyi = xi.*yi;
sxiyi = sum(xiyi);
sigmay = [syi;sxiyi];

%matrix A
A = inv(sigmax)*sigmay;
a0 = A(1);
a1 = A(2);
disp('nilai a0 =');
disp(a0);
disp('nilai a1 =');
disp(a1);
yregresi= a0 + a1*xi;
scatter(xi,yi);
hold;
plot(xi,yregresi,'b');
end

Save it into m-file with name "linierfit.m" in your default matlab folder.
Just call "linierfit(x,y)" on your command window.
As long as x (array of independent variable data) and y (array of dependent variable data) have been defined before, the call will result the value of a0 and a1 which are coefficients of linear equation best fit to the data.