Making fast informative queries with learned propagations

In an informative querying problem, one achieves a certain objective by issuing a series of queries to an oracle and receives a series of observations in return. It is a challenging task because the queries need to account for the uncertainties of the oracle, while being informative to the objective at hand. While successful algorithms have been developed for a range of querying tasks, these algorithms can be slow to compute and in some cases, intractable. A common Achilles's heel of these prior works is their reliance on the computation over the space of oracle functions itself during inference time. As a result, when the space of oracle functions becomes complex, these approaches become computationally infeasible. In this thesis, we explore an alternative approach to informative query selection. Rather than computing over the space of oracle functions, we learn a propagation function that, given a set of past observations, predicts future queries' outcomes directly. We show that by leveraging the propagation function, one can perform a range of informative querying tasks that were previously intractable. To this end, we prescribe a general method of informative querying with learned propagation: In meta-learning time, a propagation function is trained to learn the relationships between observations, and at inference time, a task specific acquisition function is constructed to leverage the propagation in making informative queries.