DeciLS-PBO: an Effective Local Search Method for Pseudo-Boolean Optimization

Luyu Jiang,Dantong Ouyang,Qi Zhang,Liming Zhang

2023-06-29

Abstract:Local search is an effective method for solving large-scale combinatorial optimization problems, and it has made remarkable progress in recent years through several subtle mechanisms. In this paper, we found two ways to improve the local search algorithms in solving Pseudo-Boolean Optimization (PBO): Firstly, some of those mechanisms such as unit propagation are merely used in solving MaxSAT before, which can be generalized to solve PBO as well; Secondly, the existing local search algorithms utilize the heuristic on variables, so-called score, to mainly guide the search. We attempt to gain more insights into the clause, as it plays the role of a middleman who builds a bridge between variables and the given formula. Hence, we first extended the combination of unit propagation-based decimation algorithm to PBO problem, giving a further generalized definition of unit clause for PBO problem, and apply it to the existing solver LS-PBO for constructing an initial assignment; then, we introduced a new heuristic on clauses, dubbed care, to set a higher priority for the clauses that are less satisfied in current iterations. Experiments on benchmarks from the most recent PB Competition, as well as three real-world application benchmarks including minimum-width confidence band, wireless sensor network optimization, and seating arrangement problems show that our algorithm DeciLS-PBO has a promising performance compared to the state-of-the-art algorithms.

Artificial Intelligence,Logic in Computer Science

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper primarily addresses the improvement of local search algorithms in Pseudo-Boolean Optimization (PBO). Specifically, the authors propose two improvement methods:

1. **Introduction of a New Heuristic Strategy**:
- A new heuristic strategy called "care" is proposed, which focuses on clauses (constraints) that are currently difficult to satisfy. Unlike traditional heuristic strategies based on variable scores, the "care" strategy emphasizes clause-level information, thereby better guiding the search direction when the algorithm falls into local optima.

2. **Extension of the Unit Propagation Algorithm**:
- The reduction algorithm based on Unit Propagation (UP) is extended for application in PBO problems to generate initial assignments. To this end, the authors define improved generalized unit clauses and make corresponding modifications to the original unit propagation algorithm.

Through the above two improvements, the paper proposes a new local search algorithm called DeciLS-PBO and validates its performance on multiple benchmarks, including Minimum Width Confidence Band (MWCB), Wireless Sensor Network Optimization (WSNO), and Seating Arrangement Problem (SAP). Experimental results show that DeciLS-PBO performs better than the existing local search algorithm LS-PBO.