Machine-learning-assisted Monte Carlo fails at sampling computationally hard problems
Simone Ciarella, Jeanne Trinquier, Martin Weigt, Francesco Zamponi
Abstract
Abstract Several strategies have been recently proposed in order to improve Monte Carlo sampling efficiency using machine learning tools. Here, we challenge these methods by considering a class of problems that are known to be exponentially hard to sample using conventional local Monte Carlo at low enough temperatures. In particular, we study the antiferromagnetic Potts model on a random graph, which reduces to the coloring of random graphs at zero temperature. We test several machine-learning-assisted Monte Carlo approaches, and we find that they all fail. Our work thus provides good benchmarks for future proposals for smart sampling algorithms.
Topics & Concepts
Monte Carlo methodComputer scienceMonte Carlo method in statistical physicsSampling (signal processing)Importance samplingMonte Carlo integrationRejection samplingHybrid Monte CarloArtificial intelligenceMachine learningAlgorithmMarkov chain Monte CarloMathematical optimizationMathematicsStatisticsFilter (signal processing)Computer visionTheoretical and Computational PhysicsMarkov Chains and Monte Carlo MethodsQuantum many-body systems