• Developed machine-translation models for predicting liquid layers over surfaces.• Designed descriptors for machine learning data pre-processing.• Conducted anomaly detection during inference and improved model performance.• Organized a machine learning and computer vision workshop for Nano in Bio Conference.

- Implemented new descriptor (LMBTR) in Dscribe package (297 ★); performed feature engineeringand model selection to predict adsorption energy of hydrogen on nanocatalysts.- Designed machine learning for material science workshop for CSC, Finland, 2018.

Developed algorithm to identify cage-like structure in super-quenched amorphous boron.

• Developed an improved Multiobjective Genetic Algorithm FORTRAN code for parametrizing potential parameters in LAMMPS MD simulation.• Conducted electronic-structure calculations using first principle, DFT based Quantum ESPRESSO package on ceramic materials.

Made phase field models of microstucture evolution during phase transformations in materials using matlab.