We are interested in correlations between parameters—in condensed matter physics and material sciences, we are trying to understand the correlation between two and more physical parameters in materials and devices, from which we optimize materials properties and improve device performance. However, establishing correlation models is often challenging due to non-linearity and the complexity of interwoven features. To uncover these correlations, numerous and systematic experiments or computations are generally required, but they are time-consuming and labor-intensive. In this presentation, I will talk about machine learning with combinatorial strategies as approaches to overcome the problems and puzzling situations mentioned above, thereby accelerating the discovery of emergent physics and quantum materials. Examples from our recent work—mainly with machine learning—will be presented.