Areas of Expertise
Deep Learning / Neural Networks, Machine Learning, Optimization algorithms, Foundation models, Python Programming
Teaching Interests
Python, Problem Solving and Programming
Research
Indra Priyadarsini is a Research Scientist at IBM Research - Tokyo, specializing in the development of models and algorithms for multimodal foundation models aimed at accelerating material discovery. Her PhD research focused on optimization algorithms and deep learning.
Representative Publications
- "SELF-BART: A Transformer-based Molecular Representation Model using SELFIES ", AI4Mat NeurIPS 2024.
- "Improving Performance Prediction of Electrolyte Formulations with Transformer-based Molecular Representation Model ”, ML4LMS ICML 2024
- "Accelerating Symmetric Rank 1 Quasi-Newton Method with Nesterov’s Gradient ”, Algorithms 2022, 15(1), 6;
- “A Nesterov’s Accelerated quasi-Newton method for Global Routing using Deep Reinforcement Learning ”, NOLTA Journal, Vol. 12(3), pp. 323-335, IEICE, Jul 2021 (Invited Paper)
Education
2019 – 2022 PhD, Shizuoka University, Japan
2017 – 2019 ME, Shizuoka University, Japan
2012 – 2016 BE, PES Institute of Technology, India
Previous Appointments
Jan 2023 – Present: Research Scientist, IBM Research – Tokyo