José E. Gallardo

On the Hybridization of Memetic Algorithms With Branch-and-Bound Techniques

Branch-and-bound (BnB) and memetic algorithms represent two very different approaches for tackling combinatorial optimization problems. However, these approaches are compatible. In this correspondence, a hybrid model that combines these two techniques is considered. To be precise, it is based on the interleaved execution of both approaches. Since the requirements of time and memory in BnB techniques are generally conflicting, a truncated exact search, namely, beam search, has opted to be carried out. Therefore, the resulting hybrid algorithm has a heuristic nature. The multidimensional 0-1 knapsack problem and the shortest common supersequence problem have been chosen as benchmarks. As will be shown, the hybrid algorithm can produce better results in both problems at the same computational cost, especially for large problem instances

Finding low autocorrelation binary sequences with memetic algorithms

This paper deals with the construction of binary sequences with low autocorrelation, a very hard problem with many practical applications. The paper analyzes several metaheuristic approaches to tackle this kind of sequences. More specifically, the paper provides an analysis of different local search strategies, used as stand-alone techniques and embedded within memetic algorithms. One of our proposals, namely a memetic algorithm endowed with a Tabu Search local searcher, performs at the state-of-the-art, as it consistently finds optimal sequences in considerably less time than previous approaches reported in the literature. Moreover, this algorithm is also able to provide new best-known solutions for large instances of the problem. In addition, a variant of this algorithm that explores only a promising subset of the whole search space (known as skew-symmetric sequences) is also analyzed. Experimental results show that this new algorithm provides new best-known solutions for very large instances of the problem.

Solving the multidimensional knapsack problem using an evolutionary algorithm hybridized with branch and bound

A hybridization of an evolutionary algorithm (EA) with the branch and bound method (B&B) is presented in this paper. Both techniques cooperate by exchanging information, namely lower bounds in the case of the EA, and partial promising solutions in the case of the B&B. The multidimensional knapsack problem has been chosen as a benchmark. To be precise, the algorithms have been tested on large problems instances from the OR-library. As it will be shown, the hybrid approach can provide high quality results, better than those obtained by the EA and the B&B on their own.

Hybridizations of Metaheuristics With Branch & Bound Derivates

An important branch of hybrid metaheuristics concerns the hybridization with branch & bound derivatives. In this chapter we present examples for two different types of hybridization. The first one concerns the use of branch & bound features within construction-based metaheuristics in order to increase their efficiancy. The second example deals with the use of a metaheuristic, in our case a memetic algorithm, in order to increase the efficiancy of branch & bound, respectively branch & bound derivatives such as beam search. The quality of the resulting hybrid techniques is demonstrated by means of the application to classical string problems: the longest common subsequence problem and the shortest common supersequence problem.

A memetic algorithm for the low autocorrelation binary sequence problem

Finding binary sequences with low auto correlation is a very hard problem with many practical applications. In this paper we analyze several meta heuristic approaches to tackle the construction of this kind of sequences. We focus on two different local search strategies, steepest descent local search (SDLS) and tabu search (TS), and their use both as stand-alone techniques and embedded within a memetic algorithm (MA). Plain evolutionary algorithms are shown to perform worse than stand-alone local search strategies. However, a MA endowed with TS turns out to be a state-of-the-art algorithm: it consistently finds optimal sequences in considerably less time than previous approaches reported in the literature.

A probabilistic beam search approach to the shortest common supersequence problem

The Shortest Common Supersequence Problem (SCSP) is a well-known hard combinatorial optimization problem that formalizes many real world problems. This paper presents a novel randomized search strategy, called probabilistic beam search (PBS), based on the hybridization between beam search and greedy constructive heuristics. PBS is competitive (and sometimes better than) previous state-of-the-art algorithms for solving the SCSP. The paper describes PBS and provides an experimental analysis (including comparisons with previous approaches) that demonstrate its usefulness.

A multi-level memetic/exact hybrid algorithm for the still life problem

Bucket elimination (BE) is an exact technique based on variable elimination. It has been recently used with encouraging results as a mechanism for recombining solutions in a memetic algorithm (MA) for the still life problem, a hard constraint optimization problem based on Conways game of life. This paper studies expanded multi-level models in which this exact/metaheuristic hybrid is further hybridized with branch-and-bound techniques. A novel variable clustering based recombination operator is also explored, with the aim of reducing the inherent time complexity of BE. Multi-parent recombination issues are analyzed as well. The obtained results are of higher quality than any previous metaheuristic approach, with large instances being solved to optimality.

Solving weighted constraint satisfaction problems with memetic/exact hybrid algorithms

A weighted constraint satisfaction problem (WCSP) is a constraint satisfaction problem in which preferences among solutions can be expressed. Bucket elimination is a complete technique commonly used to solve this kind of constraint satisfaction problem. When the memory required to apply bucket elimination is too high, a heuristic method based on it (denominated mini-buckets) can be used to calculate bounds for the optimal solution. Nevertheless, the curse of dimensionality makes these techniques impractical on large scale problems. In response to this situation, we present a memetic algorithm for WCSPs in which bucket elimination is used as a mechanism for recombining solutions, providing the best possible child from the parental set. Subsequently, a multi-level model in which this exact/metaheuristic hybrid is further hybridized with branch-and-bound techniques and mini-buckets is studied. As a case study, we have applied these algorithms to the resolution of the maximum density still life problem, a hard constraint optimization problem based on Conways game of life. The resulting algorithm consistently finds optimal patterns for up to date solved instances in less time than current approaches. Moreover, it is shown that this proposal provides new best known solutions for very large instances.

A GRASP-based memetic algorithm with path relinking for the far from most string problem

The Far From Most String Problem (FFMSP) is a string selection problem. The objective is to find a string whose distance to other strings in a certain input set is above a given threshold for as many of those strings as possible. This problem has links with some tasks in computational biology and its resolution has been shown to be very hard. We propose a memetic algorithm (MA) to tackle the FFMSP. This MA exploits a heuristic objective function for the problem and features initialization of the population via a Greedy Randomized Adaptive Search Procedure (GRASP) metaheuristic, intensive recombination via path relinking and local improvement via hill climbing. An extensive empirical evaluation using problem instances of both random and biological origin is done to assess parameter sensitivity and draw performance comparisons with other state-of-the-art techniques. The MA is shown to perform better than these latter techniques with statistical significance.

A memetic algorithm with bucket elimination for the still life problem

Bucket elimination (BE) is an exact technique based on variable elimination, commonly used for solving constraint satisfaction problems. We consider the hybridization of BE with evolutionary algorithms endowed with tabu search. The resulting memetic algorithm (MA) uses BE as a mechanism for recombining solutions, providing the best possible child from the parental set. This MA is applied to the maximum density still life problem. Experimental tests indicate that the MA provides optimal or near-optimal results at an acceptable computational cost.