Assistant Professor, UC Berkeley
Chemical Engineering and EECS
Contact info: aditik1 dot berkeley dot edu
Joining the group:
- Incoming/current UC Berkeley PhD students and prospective postdoctoral researchers: please email me directly with your research interests and CV.
- Prospective PhD students: All PhD admissions at UC Berkeley are done at the departmental level, so please apply directly to a UC Berkeley PhD program. If you apply to Chemical Engineering or EECS, you can mention my name as a faculty of interest in your application. For EECS applicants in Fall 2022, please choose CS: AI-BIO as your primary area, which in future years will be changed to CS: AI-SCIENCE.
I am interested in developing methods in machine learning that are driven by the distinct challenges and opportunities in the natural sciences, with particular interest in physics-inspired machine learning methods. Some areas of exploration include approaches to incorporate physical inductive biases into ML models to improve generalization for scientific problems, the advantages that ML can bring to classical physics-based numerical solvers (such as through end-to-end differentiable frameworks and implicit layers), and better learning strategies for distribution shifts in the physical sciences. These methods are informed by and grounded in applications in physics, fluid mechanics, molecular dynamics, materials design, climate science, and other related areas. This work also includes interfacing with other fields including numerical methods, dynamical systems theory, quantum mechanical simulations, computational geometry, and optimization. Some examples of recent work include:
Developing a differentiable PDE-constrained neural network layer to exactly enforce the relevant physics for a given problem, which can be trained end-to-end via implicit differentiation (arXiv:2207.08675; 2022),
Integrating numerical methods validation approaches for neural networks to model continuous dynamical systems (arXiv:2202.08494; 2022),
Characterizing the challenges associated with incorporating fundamental physical laws into the machine learning process (i.e., ‘‘physics-informed neural networks’’), and devising strategies to overcome their failure modes by changing the learning paradigm (Neural Information Processing Systems (NeurIPS); 2021),
Representation learning through mapping data into topological descriptors (invariant to homeomorphic transformations of the domain), for better accuracy and interpretability in renewable energy applications (structure-property relationships in nanoporous materials, proteins),
A full list of publications is available on Google scholar.