Performance Of Hyper-heuristics Research Articles

An emprical study on multi-meme memetic algorithms which includes Choice Function, Reinforcement Learning and Simple Random Hyper-heuristics *

Hyper-heuristics are designed to be reusable, domain-independent methods for addressing complex computational issues. While there are specialized approaches that work well for particular problems, they often require parameter tuning and cannot be transferred to other problems. Memetic Algorithms combine genetic algorithms and local search techniques. The evolutionary interaction of memes allows for the creation of intelligent complexes capable of solving computational problems. Hyper-heuristics are a high-level search technique that operates on a set of low-level heuristics that directly address the solution. They have two main components: heuristic selection and move acceptance mechanisms. The heuristic selection method determines which low-level heuristic to use, while the move acceptance mechanism decides whether to accept or reject the resulting solution. In this study, we explore a multi-meme memetic algorithm as a hyper-heuristic that integrates and manages multiple hyper-heuristics (Modified Choice Function All Moves, Reinforcement Learning with Great Deluge, and Simple Random Only Improvement) and parameters of heuristics (such as mutation rates and search depth). We conducted an empirical study testing two different variations of the proposed hyper-heuristic. The first algorithm uses the Only Improvement acceptance technique for both Reinforcement Learning and Simple Random, and All Moves for Modified Choice Function. In the second version, the Great Deluge method replaces Only Improvement for Reinforcement Learning. The second algorithm's results outperformed most competitors from the CHeSC2011 competition, achieving the forth-best hyper-heuristic performance.

Hyper-heuristics: A survey and taxonomy

Hyper-heuristics are search techniques for selecting, generating, and sequencing (meta)-heuristics to solve challenging optimization problems. They differ from traditional (meta)-heuristics methods, which primarily employ search space-based optimization strategies. Due to the remarkable performance of hyper-heuristics in multi-objective and machine learning-based optimization, there has been an increasing interest in this field. With a fresh perspective, our work extends the current taxonomy and presents an overview of the most significant hyper-heuristic studies of the last two decades. Four categories under which we analyze hyper-heuristics are selection hyper-heuristics (including machine learning techniques), low-level heuristics, target optimization problems, and parallel hyper-heuristics. Future research prospects, trends, and prospective fields of study are also explored.

A generality analysis of multiobjective hyper-heuristics

Selection hyper-heuristics have emerged as high level general-purpose search methodologies that mix and control a set of low-level (meta) heuristics. Previous empirical studies over a range of single objective optimisation problems have shown that the number and type of low-level (meta) heuristics used are influential to the performance of selection hyper-heuristics. In addition, move acceptance strategies play an important role and can significantly affect the overall performance of a hyper-heuristic. In this paper, we introduce an adapted variant of an existing learning automata based multiobjective hyper-heuristic from the literature. We investigate the performance and generality level of the proposed method, and another learning automata based selection hyper-heuristic, operating over a search space of multiobjective evolutionary algorithms (MOEAs) across two well-known multiobjective optimisation benchmarks. The experimental results demonstrate that, regardless of the number and type of low-level metaheuristics available, the learning automata based hyper-heuristics outperform each constituent MOEA individually, and an online learning and random choice selection hyper-heuristic from the literature. This performance and generality is shown to be consistent across a number of different move acceptance strategies.

PHH: Policy-Based Hyper-Heuristic With Reinforcement Learning

Hyper-heuristics have a high level of generality and adaptability, allowing them to effectively solve a wide range of complex optimization problems. With reinforcement learning, hyper-heuristics can use experience and knowledge gained to tackle unforeseen problems, allowing hyper-heuristics to adapt and improve over time. Our paper proposes a framework for using policy-based reinforcement learning to improve the performance of hyper-heuristics. The framework trains hyper-heuristic agents to select the best generalized constructive low-level heuristics to solve combinatorial optimization problems. The framework evaluation was performed using three benchmarking problems: traveling salesman, capacitated vehicle routing, and bin packing problems. The results showed that the proposed framework can outperform existing meta-heuristic and hyper-heuristic-based algorithms for all large problem instances in all problem domains. The proposed framework was also evaluated by applying it to a cost optimization problem for workflow scheduling on a hybrid cloud with a deadline constraint. Eight agents were trained on medium-sized workflows with two deadlines and tested against traditional meta-heuristic and hyper-heuristic methods to solve smaller and larger workflows with unforeseen deadlines. Four workflow applications, three workflow sizes, and three deadlines were used in the evaluation. The results showed that our proposed framework provided significantly better solutions of up to 98% for benchmarking problems, and up to 22% for cost optimization in workflow scheduling. Moreover, trained with small problem instances, the framework performed well for unforeseen larger problem instances implying its generalization. The proposed framework, thus, has the potential to improve both the generality and performance of solving large combinatorial optimization problems.

When move acceptance selection hyper-heuristics outperform Metropolis and elitist evolutionary algorithms and when not

Selection hyper-heuristics (HHs) are automated algorithm selection methodologies that choose between different heuristics during the optimisation process. Recently, selection HHs choosing between a collection of elitist randomised local search heuristics with different neighbourhood sizes have been shown to optimise standard unimodal benchmark functions from evolutionary computation in the optimal expected runtime achievable with the available low-level heuristics. In this paper, we extend our understanding of the performance of HHs to the domain of multimodal optimisation by considering a Move Acceptance HH (MAHH) from the literature that can switch between elitist and non-elitist heuristics during the run. In essence, MAHH is a non-elitist search heuristic that differs from other search heuristics in the source of non-elitism.We first identify the range of parameters that allow MAHH to hillclimb efficiently and prove that it can optimise the standard hillclimbing benchmark function OneMax in the best expected asymptotic time achievable by unbiased mutation-based randomised search heuristics. Afterwards, we use standard multimodal benchmark functions to highlight function characteristics where MAHH outperforms elitist evolutionary algorithms and the well-known Metropolis non-elitist algorithm by quickly escaping local optima, and ones where it does not. Since MAHH is essentially a non-elitist random local search heuristic, the paper is of independent interest to researchers in the fields of artificial intelligence and randomised search heuristics.

A Feature-Independent Hyper-Heuristic Approach for Solving the Knapsack Problem

Recent years have witnessed a growing interest in automatic learning mechanisms and applications. The concept of hyper-heuristics, algorithms that either select among existing algorithms or generate new ones, holds high relevance in this matter. Current research suggests that, under certain circumstances, hyper-heuristics outperform single heuristics when evaluated in isolation. When hyper-heuristics are selected among existing algorithms, they map problem states into suitable solvers. Unfortunately, identifying the features that accurately describe the problem state—and thus allow for a proper mapping—requires plenty of domain-specific knowledge, which is not always available. This work proposes a simple yet effective hyper-heuristic model that does not rely on problem features to produce such a mapping. The model defines a fixed sequence of heuristics that improves the solving process of knapsack problems. This research comprises an analysis of feature-independent hyper-heuristic performance under different learning conditions and different problem sets.

Assessing hyper-heuristic performance

Limited attention has been paid to assessing the generality performance of hyper-heuristics. The performance of hyper-heuristics has been predominately assessed in terms of optimality which is not ideal as the aim of hyper-heuristics is not to be competitive with state of the art approaches but rather to raise the level of generality, i.e. the ability of a technique to produce good results for different problem instances or problems rather than the best results for some instances and poor results for others. Furthermore from existing literature in this area it is evident that different hyper-heuristics aim to achieve different levels of generality and need to be assessed as such. To cater for this the paper firstly presents a new taxonomy of four different levels of generality that can be attained by a hyper-heuristic based on a survey of the literature. The paper then proposes a performance measure to assess the performance of different types of hyper-heuristics at the four levels of generality in terms of generality rather than optimality. Three case studies from the literature are used to demonstrate the application of the generality performance measure. The paper concludes by examining how the generality measure can be combined with measures of other performance criteria, such as optimality, to assess hyper-heuristic performance on more than one criterion.

Automatic Design of a Hyper-Heuristic Framework With Gene Expression Programming for Combinatorial Optimization Problems

Hyper-heuristic approaches aim to automate heuristic design in order to solve multiple problems instead of designing tailor-made methodologies for individual problems. Hyper-heuristics accomplish this through a high-level heuristic (heuristic selection mechanism and an acceptance criterion). This automates heuristic selection, deciding whether to accept or reject the returned solution. The fact that different problems, or even instances, have different landscape structures and complexity, the design of efficient high-level heuristics can have a dramatic impact on hyper-heuristic performance. In this paper, instead of using human knowledge to design the high-level heuristic, we propose a gene expression programming algorithm to automatically generate, during the instance-solving process, the high-level heuristic of the hyper-heuristic framework. The generated heuristic takes information (such as the quality of the generated solution and the improvement made) from the current problem state as input and decides which low-level heuristic should be selected and the acceptance or rejection of the resultant solution. The benefit of this framework is the ability to generate, for each instance, different high-level heuristics during the problem-solving process. Furthermore, in order to maintain solution diversity, we utilize a memory mechanism that contains a population of both high-quality and diverse solutions that is updated during the problem-solving process. The generality of the proposed hyper-heuristic is validated against six well-known combinatorial optimization problems, with very different landscapes, provided by the HyFlex software. Empirical results, comparing the proposed hyper-heuristic with state-of-the-art hyper-heuristics, conclude that the proposed hyper-heuristic generalizes well across all domains and achieves competitive, if not superior, results for several instances on all domains.

Heuristic space diversity control for improved meta-hyper-heuristic performance

This paper expands on the concept of heuristic space diversity and investigates various strategies for the management of heuristic space diversity within the context of a meta-hyper-heuristic algorithm in search of greater performance benefits. Evaluation of various strategies on a diverse set of floating-point benchmark problems shows that heuristic space diversity has a significant impact on hyper-heuristic performance. An exponentially increasing strategy (EIHH) obtained the best results. The value of a priori information about constituent algorithm performance on the benchmark set in question was also evaluated. Finally, EIHH demonstrated good performance when compared to a popular population based algorithm portfolio algorithm and the best performing constituent algorithm.

A Study of Evolutionary Algorithm Selection Hyper-Heuristics for the One-Dimensional Bin-Packing Problem

Hyper-heuristics are aimed at providing a generalized solution to optimization problems rather than producing the best result for one or more problem instances. This paper examines the use of evolutionary algorithm (EA) selection hyper-heuristics to solve the offline one-dimensional bin-packing problem. Two EA hyper-heuristics are evaluated. The first (EA-HH1) searches a heuristic space of combinations of low-level construction heuristics for bin selection. The second (EA-HH2) explores a space of combinations of both item selection and bin selection heuristic combinations. These EA hyper-heuristics use tournament selection to choose parents, and mutation and crossover with hill-climbing to create the offspring of each generation. The performance of the hyper-heuristics is compared to that of each of the low-level heuristics applied independently to solve this problem. Furthermore, the performance of both hyper-heuristics is also compared. The comparisons revealed that hyper-heuristics in general perform better than any single low-level construction heuristic in solving the problem. In addition to this it was found that the hyper-heuristic exploring a space of both item selection and bin selection heuristic combinations is more effective than the hyper-heuristic searching a space of just bin selection heuristic combinations. The performance of this hyper-heuristic was found to be comparable to other methods applied to the same benchmark sets of problems.

Monte Carlo hyper-heuristics for examination timetabling

Automating the neighbourhood selection process in an iterative approach that uses multiple heuristics is not a trivial task. Hyper-heuristics are search methodologies that not only aim to provide a general framework for solving problem instances at different difficulty levels in a given domain, but a key goal is also to extend the level of generality so that different problems from different domains can also be solved. Indeed, a major challenge is to explore how the heuristic design process might be automated. Almost all existing iterative selection hyper-heuristics performing single point search contain two successive stages; heuristic selection and move acceptance. Different operators can be used in either of the stages. Recent studies explore ways of introducing learning mechanisms into the search process for improving the performance of hyper-heuristics. In this study, a broad empirical analysis is performed comparing Monte Carlo based hyper-heuristics for solving capacitated examination timetabling problems. One of these hyper-heuristics is an approach that overlaps two stages and presents them in a single algorithmic body. A learning heuristic selection method (L) operates in harmony with a simulated annealing move acceptance method using reheating (SA) based on some shared variables. Yet, the heuristic selection and move acceptance methods can be separated as the proposed approach respects the common selection hyper-heuristic framework. The experimental results show that simulated annealing with reheating as a hyper-heuristic move acceptance method has significant potential. On the other hand, the learning hyper-heuristic using simulated annealing with reheating move acceptance (L–SA) performs poorly due to certain weaknesses, such as the choice of rewarding mechanism and the evaluation of utility values for heuristic selection as compared to some other hyper-heuristics in examination timetabling. Trials with other heuristic selection methods confirm that the best alternative for the simulated annealing with reheating move acceptance for examination timetabling is a previously proposed strategy known as the choice function.