1. [image process] build image classification CNN model on TensorFlow to search specific defect map, save labor hours 12h per case.
2. [image process] build SEM image process program to get numerical value of crystal size distribution, help to study product lifetime factors.
3. [regression analysis] analyze defect contributors and use regression model to predict defect.
4. [time series analysis] use LSTM model to predict future concentration, take early action to prevent concentration beyond control limit.
5. [process] master of dry bevel, asher, and plating process
6. [process] participate in product new generation transfer from Japan site, improve defect, yield, and product lifetime.
7. [process] publish process optimization paper and specific defect map classification CNN model paper in Micron
8. [process] maximize tool capacity to catch up product ramp up plan
9. [process] newcomer mentor

1. Study on MOF synthesis and application for CO2 adsorption and catalysis
2. Use molecular simulation RASPA package to study theoretical MOF gas adsorption behavior