• Employed machine learning and data science tools to facilitate supply chain optimization• Explored quantum computing and quantum chemistry technologies for industrial applications

Use density functional theory (DFT) and machine learning (ML) to investigate the metal-support interaction, which provides the guideline to screen through the promising candidates and design novel catalysts.• Collaborated with the statisticians in Department of Statistics at Rice University to develop new ML algorithm for feature selection that is faster and more efficient than state-of-the-art approaches.• Introduced Bayesian feature selection methods, more effective than the commonly used ML method (i.e., LASSO), to derive the predictive models for single metal binding energies on modified MgO(100).• Supervising the mentee on implementing DFT in oxide-based catalysts for propane dehydrogenationCollaboration with experimentalists in heterogeneous catalysis:• Employed DFT to reveal that the surface site arrangement contributes to 1.6 times higher in reactivity of cubic In2O3 than hexagonal In2O3 on reverse water-gas shift reaction.• Used DFT to determine the reaction mechanism for CO reduction on Cu electrocatalysts.• Discovered the nitrate reduction dependence on Pd surface facets with DFT in electrocatalysis.

Teaching assistant for Thermodynamics (Dean's TA), Transport Phenomena, and Computer Programming in Chemical Engineering

• Developed Bayesian sparse regression algorithm for linear regression and classification in Python.• Implemented Bayesian trend filtering to remove noise in time series data for interpretation and further machine learning training for forecasting.

• Studied the prospect of biotechnology industry in Taiwan.• Collected and reported the transaction data of the chemical market in Taiwan.

Designed and assembled a continuous reaction system as a pioneer to study ethanol transformation to benzene-toluene-xylene (BTX) using ZSM-5 catalyst by varying the reaction condition.