A majority of problems in engineering practice require data-driven modeling of nonlinear relationships between experimental and technological variables. The complexity of nonlinear regression techniques is gradually expanding with the development of analytical and experimental techniques; hence model structure and parameter identification is a current and important topic in the field of nonlinear regression, not just from a scientific but also from an industrial point of view as well. Model interpretability is the most important key property besides accuracy in the regression modeling of technological processes, and this is an essential characteristic of these models in their application as process controllers. As was mentioned above, model structure and parameter identification is a topic of increasing importance, since an identified model needs to be interpretable as well. In line with these expectations, and taking the interpretability of regression models as a basic requirement, robust nonlinear regression identification algorithms were developed in this book. Three algorithms were examined in detail, namely identification of regression tree based hinging hyperplanes, neural networks, and support vector regression. The application of these techniques eventuate in black-box models at the first step. We show in our book how interpretability could be maintained during model identification with utilization of applicable visualization and model structure reduction techniques within the fuzzy modeling framework.