4 ms·
You never invert A^T A anyway, you would do Cholesky on it and use it to solve A^T A x = A^T b. But regardless, QR on A should be the first choice.
by the_svd_doctor 7y ago
You never invert A^T A anyway, you would do Cholesky on it and use it to solve A^T A x = A^T b.
But regardless, QR on A should be the first choice.
- throwlaplace 7y agothat's a fair point. i would still use gradient descent (or conjugate gradients) if i didn't care about the exact minimum.