Skip to yearly menu bar Skip to main content


Accurate Quantization of Measures via Interacting Particle-based Optimization

Lantian Xu · Anna Korba · Dejan Slepcev

Hall E #701

Keywords: [ OPT: Sampling and Optimization ] [ PM: Monte Carlo and Sampling Methods ]


Approximating a target probability distribution can be cast as an optimization problem where the objective functional measures the dissimilarityto the target. This optimization can be addressed by approximating Wasserstein and related gradient flows. In practice, these are simulated by interacting particle systems, whose stationary states define an empirical measure approximating the target distribution. This approach has been popularized recently to design sampling algorithms, e.g. Stein Variational Gradient Descent, or by minimizing the Maximum Mean or Kernel Stein Discrepancy. However, little is known about quantization properties of these approaches, i.e. how well is the target approximated by a finite number particles.We investigate this question theoretically and numerically. In particular, we prove general upper bounds on the quantization error of MMD and KSD at rates which significantly outperform quantization by i.i.d. samples. We conduct experiments which show that the particle systems at study achieve fast rates in practice, and notably outperform greedy algorithms, such as kernel herding. We compare different gradient flows and highlight their quantization rates. Furthermore we introduce a Normalized Stein Variational Gradient Descent and argue in favor of adaptive kernels, which exhibit faster convergence. Finally we compare the Gaussian and Laplace kernels and argue that the Laplace kernel provides a more robust quantization.

Chat is not available.