On November 11, we organized a scientific seminar on “Towards a Fully Decentralized ALADIN Algorithm”. The scientific seminar was lead led by Kristína Fedorová, a PhD student at the Institute of Information Engineering, Automation and Mathematics FCFT STU. The scientific seminar was organised in the framework of the FrontSeat project, as part of the seminar series on “Research Seminar on Smart Cybernetics”.
Abstract:
The main benefit of the Augmented Lagrangian-based Alternating Direction Inexact Newton (ALADIN) method is the ability to solve convex and non-convex optimization problems in a distributed fashion with guaranteed convergence. However, compared to other distributed optimization algorithms, ALADIN lacks the capability of complete decentralization, which is essential for eliminating privacy concerns. Therefore, we propose a novel approach allowing the decentralization of coupled layer using the Alternating Direction Method of Multipliers (ADMM). Such an implementation provides a fully decentralized ALADIN algorithm at the cost of increased solving time.
This project has received funding from the European Union’s Horizon under grant no. 101079342 (Fostering Opportunities Towards Slovak Excellence in Advanced Control for Smart Industries).
0 Comments