On the Hardness of Probabilistic Inference Relaxations


Recent years have seen an unprecedented adoption of artificial intelligence in wide variety of applications ranging from medical diagnosis, automobile, security to aircraft collision avoidance. Probabilistic reasoning is a key component of such modern artificial intelligence systems. Sampling techniques form the core of the state of the art probabilistic reasoning systems. In contrast to testing for deterministic programs, where one trace is sufficient to prove the existence of a bug; such is not the case for samplers as one sample is typically not sufficient to prove non-conformity of the sampler to the desired distribution. This makes one wonder: whether it is possible to design testing methodology to test whether a sampler under test generates samples close to a given distribution. The primary contribution of this paper is a positive answer to the above question: We design, to the best of our knowledge, the first algorithmic framework, Barbarik, to test whether the distribution generated by $\epsilon-$close or $\eta-$far from the uniform distribution. In contrast to the sampling techniques that require an exponential or sub-exponential number of samples for sampler whose support can be represented by $n$ bits, Barbarik requires only $\mathcal{O}(1/(\eta-\epsilon)^2)$ samples. We present a prototype implementation of Barbarik and use it to test three state of the art uniform samplers over the support defined by combinatorial constraints. Barbarik is able to provide a certificate of uniformity to one sampler and demonstrate non-uniformity for the other two samplers.

In Proceedings of AAAI Conference on Artificial Intelligence (AAAI-19)