2013/11/24 by Yoshua Bengio, Bengio, Yoshua, Li Yao +3 · 3 citations
Computer Science · Physics and Astronomy · #FOS: Computer and information sciences #Gaussian Processes and Bayesian Inference #Generative Adversarial Networks and Image Synthesis #Machine Learning (cs.LG) #Model Reduction and Neural Networks #cs.LG
paper · pdf · doi:10.48550/arxiv.1311.6184
10 pages, 1 figure, 2 tables. International Conference on Learning Representations (ICLR'2014, conference track)
openalex publication_date 2013/11/24 · arxiv created 2014/05/09 · arxiv updated 2014/05/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Several interesting generative learning algorithms involve a complex probability distribution over many random variables, involving intractable normalization constants or latent variable normalization. Some of them may even not have an analytic expression for the unnormalized probability function and no tractable approximation. This makes it difficult to estimate the quality of these models, once they have been trained, or to monitor their quality (e.g. for early stopping) while training. A previously proposed method is based on constructing a non-parametric density estimator of the model's probability function from samples generated by the model. We revisit this idea, propose a more efficient estimator, and prove that it provides a lower bound on the true test log-likelihood, and an unbiased estimator as the number of generated samples goes to infinity, although one that incorporates the effect of poor mixing. We further propose a biased variant of the estimator that can be used reliably with a finite number of samples for the purpose of model comparison.