For a demand to accelerate materials development, a new trend research which focusses on “integration” has been received much attention, where informatics approach tends to be incorporated into material science. Depending on a subject, we are required to select a machine learning method such as artificial neural network(ANN), support vector machine, Bayesian inference, convolutional ANN (deep learning) et al.  3D4D characterization is beneficial to obtain fruitful materials genome, i.e. metric features and topological features such as connectivity. However, it is time-consuming to capture a 3D microstructure image followed by image processing and quantification. Subsequently descriptors of materials genome are input into modelling or machine learning to estimate a property. Therefore, many materials scientists likely hope to accelerate all these processes. This laboratory offeres both a fully automated serial sectioning 3D microscope “Genus_3D” and material genome integration system for phase and property analysis “MIPHA”. MIPHA consists of microstructure recognition module by deep learning, microstructure detection module, 2D analysis module, 3D analysis module, property estimation module, and inverse analysis module. Various kinds of machine learning methods are implemented in these modules.