Welcome!

 
Thank you for visiting my personal page. I am an Associate Professor at The Institute of Statistical Mathematics, Tokyo, Japan. I was a visiting researcher at the National Institute for Materials Science, Tsukuba, Japan. My primary research interest focuses on machine learning and Bayesian modeling for a wide range of applications, including materials informatics, bioinformatics, pharmacokinetics, molecular dynamics and earthquake engineering. My current research focus is on Bayesian machine learning for materials informatics. In particular, I am one of the active developers of XenonPy, which is an all-in-one open source package for material design.

I received my B.S. degree in Civil & Environmental Engineering and Mathematics from University of Michigan - Ann Arbor in 2008. My undergraduate research advisor was Prof. Jerome P. Lynch. Afterward, I received my M.S. and PhD degree from the Department of Mechanical and Civil Engineering at the California Institute of Technology (Caltech). I was studying probability based earthquake engineering problems in Prof. James L. Beck's research group. My thesis is on engineering applications of Earthquake Early Warning (EEW) system. After my PhD graduation, I have researched on hierarchical Bayesian models for molecular dynamics and pharmacokinetics problems as a postdoc at the Computational Science & Engineering Laboratory (CSELab), ETH-Zurich, Switzerland under the guidance of Prof. Petros Koumoutsakos, data assimilation and statistical modeling of C. elegans neural network as a project assistant professor at my current institute under Prof. Ryo Yoshida's supervision.

My research involves uses and developments of many machine learning and Monte Carlo sampling algorithms, such as, neural network transfer learning, GAN-based inverse problem solving, Subset Simulation on Complex Network problems, development of stochastic optimization techniques based on sparse Bayesian learning algorithm (e.g. relevance vector machine - RVM) on Compressive Sensing problems, Bayesian optimization with Gaussian Process and value of information for Geotechnical applications, parallel and unbiased Transitional Markov Chain Monte Carlo (e.g. BASIS) for hierarchical Bayesian modeling, principal component analysis and coherent motion registration for pan-neuronal imaging, etc.  

Recent News:

Last updated: July 28, 2020

 

New publication: Huang et al. (2020), Recovering compressed images for automatic crack segmentation using generative models.

New publication: Wu et al. (2019), iQSPR in XenonPy: A Bayesian Molecular Design Algorithm, Molecular Informatics.

New publication: Yamada&Liu et al. (2019), Predicting Materials Properties with Little Data Using Shotgun Transfer Learning, ACS Central Science.

2018 Best paper award: Wu et al. (2017), Bayesian Annealed Sequential Importance Sampling (BASIS): an unbiased version of Transitional Markov Chain Monte Carlo, ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part B: Mechanical Engineering, 4(1):011008.

New publication: Wu et al. (2019), Machine-learning-assisted discovery of polymers with high thermal conductivity using a molecular design algorithm, npj Computational Materials.