معرفی
Manu Upadhyaya is a Doctoral Candidate at Lund University's Department of Automatic Control within the Faculty of Engineering. His research focuses on continuous optimization with applications in machine learning, control systems, and finance. He is actively involved in the Department of Automatic Control's research ecosystem and collaborates internationally with institutions including Inria Paris, UC Berkeley, and ETH Zürich.
His educational background includes an MSc in Engineering Physics (2020), an MSc in Finance (2020), and a BSc in Mathematics (2015), all from Lund University. His PhD defense is scheduled for September 26, 2025, under supervisors Pontus Giselsson and Sebastian Banert.
Upadhyaya specializes in the design and performance analysis of first-order algorithms for convex optimization problems. His recent work centers on automated Lyapunov analysis frameworks for verifying convergence properties of optimization algorithms, bridging theoretical guarantees with practical implementation. This research has direct applications in machine learning systems and control theory where computational efficiency is critical.
His publications demonstrate a strong trend toward computer-aided analysis of optimization algorithms, with emphasis on tight convergence guarantees and automated verification methods. Key contributions appear in high-impact venues like Mathematical Programming, focusing on systematic approaches to algorithm design.
- Co-organizer of EUROPT 2024 (21st Conference on Advances in Continuous Optimization)
- Active participant in WASP (Wallenberg AI, Autonomous Systems and Software Program)
- Member of ELLIIT (Excellence Center at Linköping–Lund in Information Technology)
Upadhyaya leads research projects including 'Automatic Lyapunov analysis of optimization algorithms' and contributed to ESA-funded 'AI4GNC' (Artificial intelligence for guidance, navigation, and control). His work connects theoretical optimization with real-world applications in cloud computing and autonomous systems through Lund's RobotLab and WASP initiatives. Current projects focus on scalable control of interconnected systems and optimization for learning frameworks.
