Optimization Framework

Member： Jean-François Baffier

Category: Innovating

Tags: optimization, user-friendly, AI, learning

1. Background：Efficient, adaptive optimization tools are essential for complex resource allocation and traffic engineering, necessitating a balance between efficiency and modeling ease.
2. Purpose：To simplify complex decision-making through a semi-automated optimization framework that doesn’t compromise efficiency.

Recent Progress

• Semi-automated Framework: Achieved a user-friendly solution for non-experts and customizable options for experts.​

• ConstraintsLearning.jl: Developed a library for both classical and quantum devices, integrating the XCSP3-core standard for constraint learning.​

• Performance Tools: Launched PerfChecker.jl for clean-environment performance checks and Continuous Integration support, enhancing tool development.