Postdoctoral Scholar

The Friedman School of Nutrition Science and Policy; Medford (MA); Hybrid; USD 61,000 - 74,000; Full time

Job summary

The Friedman School of Nutrition Science and Policy at Tufts University invites applications for a Postdoctoral Scholar in Machine Learning for Physical Systems. The role focuses on developing ML emulators and surrogate models for chaotic dynamics and materials modeling, with opportunities to shape the research alongside the PI.

Responsibilities include independent and collaborative research, implementing and validating scientific ML models, publishing results, and contributing to grant

Qualifications

• PhD in electrical and computer engineering, physics, applied mathematics, computer science, or closely related field.
• Strong record of research and ability to work independently.
• Proficiency in scientific computing and modern ML frameworks.
• Preferred: PDEs, dynamical systems, turbulence, materials modeling, or ML surrogate models; differentiable programming.

Responsibilities

• Conduct independent and collaborative research in ML for physical systems.
• Develop and validate scientific ML models and code.
• Disseminate results through publications and presentations; contribute to grant activity.
• Mentor student researchers as part of the project.

Skills

ML frameworks

Scientific computing

Dynamical systems

PDEs experience

Turbulence modeling

Materials modeling

ML surrogate models

Differentiable programming

Research excellence

Python

Education

PhD (completed/expected)

Tools

Python

PyTorch

TensorFlow

Job description

Overview

Postdoctoral Scholar in Machine Learning for Physical Systems

The Department of Electrical and Computer Engineering at Tufts University invites applications for a Postdoctoral Scholar in the research group of Prof. Peter Lu. The position focuses on machine learning methods for physical systems, including ML emulators or surrogate models for chaotic dynamics and materials modeling. The successful candidate will develop and analyze ML models for scientific problems and will have latitude to shape the research direction in collaboration with the PI

What You'll Do

Relevant application domains include high-dimensional PDEs, turbulence, and materials modeling. Methodological interests in the group span scientific generative modeling, representation learning, optimal transport, and neural operators. Candidates whose expertise connects to any of these areas, and whowant to work at the interface of rigorous physical modeling and modern ML are encouraged to apply.

Responsibilities include conducting independent and collaborative research, developing and validating scientific ML models and code, disseminating results through publications and presentations, and contributing to grant activity and the mentoring of student researchers.

What We're Looking For

Required qualifications: a PhD (completed or expected before the start date) in electrical and computer engineering, physics, applied mathematics, computer science, or a closely related field; a strong record of research; and proficiency in scientific computing and modern ML frameworks.

Preferred qualifications: demonstrated experience in one or more of PDEs, dynamical systems, turbulence, materials modeling, or ML surrogatemodels; familiarity with differentiable programming; and a background spanning both physical modeling and machine learning.

The preferred start date is as soon as possible. The appointment is for up to two years: an initial one-year term with a second year contingent on performance and funding. Salary follows the Tufts compensation schedule for postdoctoral scholars and is commensurate with experience.

Pay Range

Minimum $67,500.00, Midpoint $67,500.00, Maximum $67,500.00

Salary is based on related experience, expertise, and internal equity; generally, new hires can expect pay between the minimum and midpoint of the range.