
Indie researcher drops Hvala algorithm: crushes vertex cover on 233 graphs at 1.071 ratio, casually claims it might solve P=NP—if you squint
Researchers at the Information Physics Institute have developed the Hvala algorithm, an ensemble approximation method for the Minimum Vertex Cover problem. Led by Frank Vega, the study presents a novel reduction technique, optimal solvers on reduced graph structures, and complementary heuristics. Tested on 233+ diverse instances, the algorithm achieved consistent approximation ratios between 1.001 and 1.071, with no instance exceeding 1.071. Theoretical analysis proves optimality on specific graph classes, including paths, trees, and complete graphs. The Hvala algorithm operates in O(mlogn) time and O(m) space, making it efficient for large-scale graphs. While the study's hypothesis that the algorithm achieves a provable approximation ratio less than 2 is intriguing, it remains to be proven, and the research community is invited to verify or refute the claim. The algorithm is publicly available on PyPI, and source code is provided for independent verification. If proven, the hypothesis would imply P = NP, solving one of the seven Millennium Prize Problems.