Use the polyder function to get the first derivative (and, if you want to be efficient, the second derivative as well), then use the roots function and basic calculus to find the minimum.
Example:
p = polyfit([-1:0.01:2], cos(-1:0.01:2),6); % Polynomial Coefficients
d1p = polyder(p); % First Derivative
d2p = polyder(d1p); % Second Derivative
ips = roots(d1p); % Inflection Points
xtr = polyval(d2p, ips); % Evaluate ‘d2p’ at ‘ips’
minpts = ips((xtr > 0) & (imag(xtr)==0)); % Find Minima
x = linspace(min(ips)-5,max(ips)+5);
ep = polyval(p,x);
figure(1)
plot(x, ep, '-r')
hold on
plot(minpts, polyval(p,minpts), 'bp')
hold off
grid
