Abstract: In this seminar, I will briefly introduce the Department of Materials Science and Engineering at KAIST and present two main themes: visualization of polarization using atomic force microscopy and machine-learning-based materials design and development.

In the first part, I will show how proxies like piezoelectric strain, screening charges, and friction coefficient can be used to image ferroelectric polarization and how the contrast mechanism can be applied to energy harvesting and chemical mechanical polishing. In the second part, I will show how machine learning can help us find the quantitative correlation between processing, structure, measurement, and performance parameters from related journal papers and how it can be used to develop processing recipes for improved performance. Materials imaging combined with machine learning will be compelling shortly and accelerate the speed of analysis without sacrificing accuracy, leading to more competitive failure analysis and new materials design. Finally, I will briefly cover the materials and molecular modeling, imaging, informatics, and integration (M3I3) initiative at KAIST and conclude my talk.