The Kernel-based Cumulative Sum (KCUSUM) algorithm is a non-parametric extension of the traditional Cumulative Sum (CUSUM) method, which can effectively detect changes in real-time data streams without requiring prior knowledge of the underlying data distribution.
The paper presents an improved approximation algorithm for the Set Family Edge Cover problem with pliable set families, achieving an approximation ratio of 10, which improves upon the previous ratio of 16.
This article presents the first algorithm that runs in O(n log n) time in the worst case for the Signed Sorting by Reversals problem, which transforms a signed permutation into the identity permutation using a minimum-length sequence of reversals.
Given n positive integers with total sum less than 2^n-1, the algorithm efficiently finds two distinct subsets with equal subset sums.
Prior diffusion in Langevin algorithms enables dimension-independent convergence for non-log-concave distributions.
FirstFit and Online-EDF algorithms analyzed for Interval-Constrained Bipartite Matching.
Predictions can be leveraged to improve Max-Cut approximation ratios, with ε-accurate predictions enhancing algorithm performance.
Efficient algorithms reduce computing plans between junctions to two problems: optimal energetic paths and standard shortest paths.
Developing a versatile greedy-based technique for Single-Sample Prophet Inequalities directly, improving competitive guarantees.
Greedy algorithms outperform UCB in many-armed bandit problems.