Research Fellow (Relative smooth optimization theory)

NATIONAL UNIVERSITY OF SINGAPORE

Singapore, Singapore
Not specified (assumed on-site due to university setting)
Phd in 2026
Convergence and complexity analysis for nonconvex optimization
Variational inequalities and duality theory
The National University of Singapore is seeking a Research Fellow in relative smooth optimization theory, focusing on enhancing the theoretical foundations of optimization methods applicable to complex, nonconvex problems. The ideal candidate should possess a PhD by 2026 and have expertise in optimization algorithms, coding in MATLAB and Python, and experience with GPU-based algorithm acceleration

Job Summary

  • This project aims to strengthen the theoretical foundation of relative smooth optimization by studying first-order methods under relative smoothness.
  • The role focuses on establishing sharper convergence and complexity results for nonconvex problems using new non-Euclidean Lipschitz tools.
  • Candidates must possess a PhD in 2026 with strong expertise in variational inequalities, stochastic processes, and dynamical systems.

Matching Summary

Match Score: 80

The National University of Singapore is seeking a Research Fellow in relative smooth optimization theory, focusing on enhancing the theoretical foundations of optimization methods applicable to complex, nonconvex problems. The ideal candidate should possess a PhD by 2026 and have expertise in optimization algorithms, coding in MATLAB and Python, and experience with GPU-based algorithm acceleration.

Skills & Requirements

Must-have

  • PhD in 2026
  • convergence and complexity analysis for nonconvex optimization
  • variational inequalities and duality theory
  • stochastic process and martingale theory
  • semi-algebraic and subanalytic geometry
  • stochastic approximation methods
  • dynamical systems
  • MATLAB and Python coding
  • PyTorch code adaptation

Nice-to-have

  • GPU-based acceleration of large-scale algorithms
  • implementing first-order methods on GPU platforms
  • handling large-scale matrix-vector computations efficiently

Key Requirements

  • PhD degree expected in 2026
  • Experience adapting PyTorch base codes
  • Proficiency in MATLAB and Python

Work Rights

Not specified

Tailored Resume

Cover Letter