🤖 AI Summary
This work addresses deterministic graph algorithms in the synchronous message-passing model, focusing on leveraging potentially erroneous predictions to accelerate computation while ensuring robustness when predictions fail. We introduce the first prediction-augmented paradigm tailored to distributed graph algorithms, featuring a graph-structure-aware error metric and a generic construction template that achieves provably optimal trade-offs between speedup and robustness. Our method integrates prediction error modeling, hierarchical activation, and local adaptive decision-making, and applies to fundamental problems including maximal independent set (MIS), maximum matching (MM), (Δ+1)-vertex coloring, and (2Δ−1)-edge coloring. Notably, we provide theoretically optimal prediction-augmented algorithms for MIS on general graphs and rooted trees. The framework establishes a new theoretical foundation for distributed graph learning and prediction-driven systems, delivering both rigorous guarantees and practical algorithmic tools.
📝 Abstract
We initiate the study of deterministic distributed graph algorithms with predictions in synchronous message passing systems. The process at each node in the graph is given a prediction, which is some extra information about the problem instance that may be incorrect. The processes may use the predictions to help them solve the problem. The overall goal is to develop algorithms that both work faster when predictions are good and do not work much worse than algorithms without predictions when predictions are bad. Concepts from the more general area of algorithms with predictions, such as error measures, consistency, robustness, and smoothness, are adapted to distributed graph algorithms with predictions. We consider algorithms with predictions for four distributed graph problems, Maximal Independent Set, Maximal Matching, $(Delta+1)$-Vertex Coloring, and $(2Delta-1)$-Edge Coloring, where $Delta$ denotes the degree of the graph. For each, we define an appropriate error measure. We present generic templates that can be used to design deterministic distributed graph algorithms with predictions from existing algorithms without predictions. Using these templates, we develop algorithms with predictions for Maximal Independent Set. Alternative error measures for the Maximal Independent Set problem are also considered. We obtain algorithms with predictions for general graphs and for rooted trees and analyze them using two of these error measures.