Kehang

Hi there, thanks for visiting my website!


I'm currently a PhD student at MIT with research focused on high performance computing, machine learning and energy. I love building models and making predictions, enjoy doing self-organized fun projects with friends (like this one).

A little background about myself. I was born and grew up in southern China, later moved north for college (Tsinghua Univ.) where I participated in lots of research projects from catalyst, polymer, protein to finally optimization algorithm design. In 2012 I moved again to US for graduate education in MIT.

Now I’m the lead developer of an open source project RMG-Py, an automatic reaction network generation software suite developed in Green Group at MIT. Over the years, it's gaining large user base including many universities, institutions and companies e.g., ExxonMobil, Eni, Department of Energy.

My research interest is to apply high performance computing and machine learning techniques into RMG for computational energy applications. I’ve been working on

  • RMG memory usage optimization and parallelism
  • molecular property predictions by convolutional neural networks
  • construction of integration test platform for RMG
  • subsurface oil and gas modeling for Eni

  • In my spare time, I like to do fun projects which include

  • Smart Amazon Review Summarizer
  • Sidney-Pacific website chair
  • Hand Gesture Recognition
  • Web version of RMG-Py
  • Data-driven energy disaggregator  (3rd prize in MIT Energy Hackathon)

  • For more details, please visit my project page and blog.