On the usefulness of linkage processing for solving MAX-SAT

@inproceedings{Sadowski2013OnTU,
  title={On the usefulness of linkage processing for solving MAX-SAT},
  author={Krzysztof L. Sadowski and Peter A. N. Bosman and Dirk Thierens},
  booktitle={GECCO '13},
  year={2013}
}
Mixing of partial solutions is a key mechanism used for creating new solutions in many Genetic Algorithms (GAs). However, this mixing can be disruptive and generate improved solutions inefficiently. Exploring a problem's structure can help in establishing less disruptive operators, leading to more efficient mixing. One way of using a problem's structure is to consider variable linkage information. Once a proper linkage model for a problem is obtained, mixing becomes more efficient. This paper… 

Figures and Tables from this paper

Parallelizing the Linkage Tree Genetic Algorithm and Searching for the Optimal Replacement for the Linkage Tree

TLDR
Two parallel implementations of LTGA are presented that enable us to leverage the computational power of a multi-processor architecture and solve a problem that previously could not be solved, being the problem of finding high-quality predetermined linkage models that result in a better performance ofLTGA for intricate problems by replacing the online-learned LTs.

Empirical performance evaluation of the linkage tree genetic algorithm

TLDR
It is concluded that certain problem instances tested, Low Autocorrelation Binary Sequences, two instances of the Ising model, and the Maximum Independent Vertex Set benefit from linkage information as the LTGA performed very good compared to other algorithms on these.

A Model-based Genetic Programming Approach for Symbolic Regression of Small Expressions

TLDR
The role of Linkage Learning (LL) performed by GOMEA in Symbolic Regression (SR) is studied, and it is shown that the non-uniformity in the distribution of the genotype in GP populations negatively biases LL, and a method to correct this is proposed.

Optimization by Pairwise Linkage Detection, Incremental Linkage Set, and Restricted / Back Mixing: DSMGA-II

TLDR
Experimental results show that DSMGA-II outperforms LT-GOMEA and hBOA in terms of number of function evaluations on the concatenated/folded/cyclic trap problems, NK-landscape problems with various degrees of overlapping, 2D Ising spin-glass problems, and MAX-SAT.

Improving Model-Based Genetic Programming for Symbolic Regression of Small Expressions

TLDR
This article shows that the non-uniformity in the distribution of the genotype in GP populations negatively biases LL, and proposes a method to correct this, and finds that GOMEA is a promising new approach to SR.

Review on Adaptive Genetic Algorithm and Metaheuristic Methods within Stochastic Optimisation

In genetic algorithm each candidate solution has a set of properties (its chromosomes or genotype) which can be mutated and altered; traditionally, solutions are represented in binary as strings of

A Stochastic Local Search Algorithm for the Partial Max-SAT Problem Based on Adaptive Tuning and Variable Depth Neighborhood Search

TLDR
A novel SLS algorithm called adaptive variable depth SLS for PMSAT problem solving based on a dynamic local search framework is proposed and results show that AVD-SLS solver, which is implemented based on the algorithm, outperforms state-of-the-art PMSat SLS solvers in most benchmark classes, including random, crafted, and industrial instances.

A diversity preservation scheme for DSMGA-II to conquer the hierarchical difficulty

TLDR
A diversity preservation scheme for DSMGA-II to conquer the hierarchical difficulty by calculating the entropies of subsolutions and determining whether to perform the back mixing and the empirical results show that the algorithm works well on hierarchical problems and does not compromise the performance on other problems.

Empirical investigation of stochastic local search for maximum satisfiability

TLDR
Experimental results show that the enhanced SLS algorithm developed here performs better than its state-of-the-art SLS competitors on a large number of industrial MAX-SAT instances.

References

SHOWING 1-10 OF 13 REFERENCES

Linkage neighbors, optimal mixing and forced improvements in genetic algorithms

TLDR
This paper considers a technique called Forced Improvements (FI), that allows LTGA to converge to a single solution without requiring an explicit, diversity-reducing, selection step, and considers a different linkage model, called Linkage Neighbors (LN), that is more flexible, yet can be learned equally efficiently from data.

Predetermined versus learned linkage models

TLDR
Experimental results show that - for problems with intricate interaction structure - the a priori fixed models are actually less efficient than LTGA that dynamically learns a hierarchical tree model.

The Linkage Tree Genetic Algorithm

TLDR
Experimental results for fully deceptive functions and nearest neighbor NK-landscape problems with tunable overlap show that the LTGA can solve these hard functions efficiently without knowing the actual position of the linked variables on the problem representation.

Optimal mixing evolutionary algorithms

TLDR
The efficiency of mixing in genetic algorithms (GAs) and estimation-of-distribution algorithms (EDAs) is looked at and two new Evolutionary Algorithms that maximize the juxtaposing of the partial solutions present in the parents are proposed.

BOA: the Bayesian optimization algorithm

TLDR
Preliminary experiments show that the BOA outperforms the simple genetic algorithm even on decomposable functions with tight building blocks as a problem size grows.

A New Method for Solving Hard Satisfiability Problems

TLDR
A greedy local search procedure called GSAT is introduced for solving propositional satisfiability problems and its good performance suggests that it may be advantageous to reformulate reasoning tasks that have traditionally been viewed as theorem-proving problems as model-finding tasks.

Maximum Satisfiability: Anatomy of the Fitness Landscape for a Hard Combinatorial Optimization Problem

TLDR
These analyses show why the problem becomes hard for local search algorithms as the system size increases and how a recently proposed algorithm can exploit long-range correlations in the fitness landscape to improve on the state-of-the-art heuristic algorithms.

Genetic Algorithm Behavior in the MAXSAT Domain

TLDR
This work has shown that all nonzero Walsh coefficients can be computed exactly for MAXSAT problems in polynomial time using only the clause information, and unless P=NP, this low order information cannot reliably lead to the global optimum, thus nontrivial MAXSat problems must be deceptive.

Local search strategies for satisfiability testing

TLDR
The power of local search for satissability testing can be further enhanced by employing a new strategy, called mixed random walk, for escaping from local minima, which allows us to handle formulas that are substantially larger than those that can be solved with basic local search.

SATLIB: An Online Resource for Research on SAT

TLDR
An overview of SATLIB is given and its current set of benchmark problems are described, to encourage all members of the community to utilise it for their SAT-related research and to improve it by submitting new benchmark problems, SAT solvers, and bibliography entries.