Transfer Learning of Surrogate Models: Integrating Domain Warping and Affine Transformations