Brief

Researchers have developed algorithms to approximate the total variation distance between mixtures of product distributions. The work focuses on an n-dimensional discrete domain and provides a randomized algorithm for approximation within a $(1 \pm \varepsilon)$ error. For mixtures of Boolean subcubes, a deterministic algorithm offers exact computation, though the problem is shown to be #P-hard under certain conditions. AI

Researchers have developed a new data structure for the incremental strongly connected components (SCC) problem, which involves maintaining the SCCs of a directed graph as edges are added over time. This algorithm leverages machine-learned predictions about the edge sequence to precompute partial solutions, aiming for faster insertions. The theoretical analysis shows that the algorithm achieves nearly optimal bounds with accurate predictions, and its performance degrades gracefully with prediction errors. Experimental results on real datasets indicate that the theoretical predictions align with practical runtime improvements. AI

Researchers have introduced a new cost model for matroid algorithms that accounts for the size of queried sets, moving beyond the traditional constant-time assumption. This size-sensitive approach better reflects the actual computational effort, particularly for natural matroid classes like graphic matroids. The study establishes tight bounds for fundamental tasks such as finding a basis and approximating rank, showing optimal query costs are generally quadratic in the matroid size, with exceptions for matroids with small maximum circuit sizes. AI