Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This lecture introduces the Random-Subcube Model (RSM) as a toy model for constraint satisfaction problems. The RSM reproduces the structure of solution spaces in random k-satisfiability and k-coloring problems, undergoing similar phase transitions. The model's parameters a and p play crucial roles in defining frozen and free variables within clusters, with a representing constraint density and p determining variable freezing probability. The lecture delves into the RSM's properties, such as internal entropy, cluster distribution, and phase transitions, shedding light on its significance in understanding combinatorial optimization problems.