Two points is not enough to fit a cubic. You'd need at least 4 points. If you have 4 x and 4 y then you can do
x = sort(10 * rand(1, 4));
y = 10 * rand(1, 4);
plot(x, y, 'b.', 'MarkerSize', 30);
grid on;
coefficients = polyfit(x, y, 3)
xFit = linspace(min(x), max(x), 1000);
yFit = polyval(coefficients, xFit);
hold on;
plot(xFit, yFit, 'r-', 'lineWidth', 2);
xlabel('x');
ylabel('y');
legend('training data', 'Fitted Curve')
If you have any more questions, then attach your data and code to read it in with the paperclip icon after you read this:
