Postdoctoral Research Associate in Causal AI

American Medical Informatics Association, Inc; Indiana; In loco; USD 48.000 - 60.000; Tempo pieno

Descrizione del lavoro

Purdue University invites applications for a Postdoctoral Research Associate to develop a causal multimodal foundation model using EHR trajectories, molecular profiles, and wearable data. The fellow will work under co-mentors Nan Kong and Jingchuan Serena Guo, contributing to representation learning, causal inference, and longitudinal analysis.

The role emphasizes interdisciplinary collaboration, mentorship, and professional development to support an independent academic career.

Competenze

• PhD completed or near completion by start date.
• Strong quantitative and computational skills with experience in causal inference and ML.

Mansioni

• Develop and evaluate new methods combining multimodal representation learning and causal treatment-effect estimation.
• Integrate and analyze biomedical data; assess transferability, uncertainty, and subgroup responses.
• Lead manuscripts, contribute to grant proposals, and collaborate across engineering, data science, and health sciences.

Conoscenze

Causal inference

Multimodal learning

Longitudinal modeling

Optimization

Biomedical data analytics

Python

Machine learning frameworks

Formazione

PhD in operations research, industrial or biomedical engineering, biomedical/health informatics, computer science, statistics, biostatistics, computational biology, or related field

Strumenti

Python

PyTorch

TensorFlow

Descrizione del lavoro

Applications are invited for a Postdoctoral Research Associate position at Purdue University, co-mentored by Nan Kong, PhD (Biomedical Engineering and Industrial Engineering), and Serena Guo, MD, PhD (Pharmacy Practice). The project will develop a causal multimodal foundation model integrating electronic health record (EHR) trajectories, molecular profiles, and longitudinal wearable/mobile sensor data to estimate treatment effects and identify heterogeneous treatment responses.

The fellow will combine representation learning, causal inference, longitudinal modeling, and operations research, with interdisciplinary mentorship and professional development to support an independent academic career.

Research Areas:

The fellow will contribute to the following interconnected areas:

1. Multimodal foundation models and transferable patient representations across temporal scales, missingness patterns, and measurement processes
2. Causal inference for longitudinal treatment effects, addressing confounding and clinical treatment selection
3. Treatment-response heterogeneity and operations research methods for individualized treatment decision-making

Primary Responsibilities:

1. Develop and evaluate new methods combining multimodal representation learning and causal treatment-effect estimation.
2. Integrate and analyze biomedical data; assess model transferability, uncertainty, and subgroup treatment responses.
3. Lead manuscripts, contribute to grant proposals and reproducible research tools, and collaborate across engineering, data science, and health sciences.

Application Materials:

Curriculum vitae; cover letter describing research interests, relevant methodological and computational experience, and career goals; and contact information for three references.

Co-Mentors / Application Contacts:

1. Nan Kong, PhD — Professor of Biomedical Engineering and Industrial Engineering; [e-mail]
2. Jingchuan Serena Guo, MD, PhD — Associate Professor of Pharmacy Practice; [e-mail]

1. PhD in operations research, industrial or biomedical engineering, biomedical/health informatics, computer science, statistics, biostatistics, computational biology, or a related field. Candidates nearing completion of their doctorate are welcome to apply; the degree must be completed by the appointment start date.
2. Strong quantitative and computational skills, with experience in one or more of causal inference, machine learning/foundation models, multimodal learning, longitudinal modeling, optimization, or biomedical data analytics.
3. Demonstrated methodological research ability, scholarly productivity, and strong communication and collaboration skills. Python proficiency and experience with modern machine-learning frameworks are desirable.

Additional Fellowship Opportunity:

Strong candidates may also have the opportunity to apply for Purdue’s Lillian Gilbreth Postdoctoral Fellowships, which offer interdisciplinary research training and professional development for academic careers. Fellowship awards are subject to program eligibility and competitive selection.