Search papers, labs, and topics across Lattice.
This paper introduces a method for compressing lifted probability distributions by extracting first-order logic formulas representing regions of the distribution. The goal is to reduce the exponential space requirements of tabular distribution representations while preserving key information. Experiments demonstrate that this approach can significantly increase sparsity by representing distributions with a small set of short formulas.
Representing probability distributions with first-order logic formulas can drastically reduce their size, offering a path to more efficient probabilistic reasoning.
Real world scenarios can be captured with lifted probability distributions. However, distributions are usually encoded in a table or list, requiring an exponential number of values. Hence, we propose a method for extracting first-order formulas from probability distributions that require significantly less values by reducing the number of values in a distribution and then extracting, for each value, a logical formula to be further minimized. This reduction and minimization allows for increasing the sparsity in the encoding while also generalizing a given distribution. Our evaluation shows that sparsity can increase immensely by extracting a small set of short formulas while preserving core information.
