Density Ratio Estimation via Infinitesimal Classification
arXiv:2111.11010
Abstract
Density ratio estimation (DRE) is a fundamental machine learning technique for comparing two probability distributions. However, existing methods struggle in high-dimensional settings, as it is difficult to accurately compare probability distributions based on finite samples. In this work we propose DRE-\infty, a divide-and-conquer approach to reduce DRE to a series of easier subproblems. Inspired by Monte Carlo methods, we smoothly interpolate between the two distributions via an infinite continuum of intermediate bridge distributions. We then estimate the instantaneous rate of change of the bridge distributions indexed by time (the "time score") -- a quantity defined analogously to data (Stein) scores -- with a novel time score matching objective. Crucially, the learned time scores can then be integrated to compute the desired density ratio. In addition, we show that traditional (Stein) scores can be used to obtain integration paths that connect regions of high density in both distributions, improving performance in practice. Empirically, we demonstrate that our approach performs well on downstream tasks such as mutual information estimation and energy-based modeling on complex, high-dimensional datasets.
arXiv commentFirst two authors contributed equally
References in corpus (16)
- Denoising Diffusion Probabilistic Models
- Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
- Generative Modeling by Estimating Gradients of the Data Distribution
- Searching for Activation Functions
- f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
- Improved Denoising Diffusion Probabilistic Models
- Variational Diffusion Models
- On Variational Bounds of Mutual Information
- Improved Techniques for Training Score-Based Generative Models
- Sliced Score Matching: A Scalable Approach to Density and Score Estimation
- NeuTra-lizing Bad Geometry in Hamiltonian Monte Carlo Using Neural Transport
- Maximum Likelihood Training of Score-Based Diffusion Models
- Learning Weighted Representations for Generalization Across Designs
- Relative Density-Ratio Estimation for Robust Distribution Comparison
- Adaptive Path Sampling in Metastable Posterior Distributions
- Featurized Density Ratio Estimation