Optimisation and learning

Evaluating the Robustness of Deep-Learning Algorithm-Selection Models by Evolving Adversarial Instances (2024)
Presentation / Conference Contribution
Hart, E., Renau, Q., Sim, K., & Alissa, M. (2024, September). Evaluating the Robustness of Deep-Learning Algorithm-Selection Models by Evolving Adversarial Instances. Presented at 18th International Conference on Parallel Problem Solving From Nature PPSN 2024, Hagenburg, Austria

Deep neural networks (DNN) are increasingly being used to perform algorithm-selection in combinatorial optimisation domains, particularly as they accommodate input representations which avoid designing and calculating features. Mounting evidence fro... Read More about Evaluating the Robustness of Deep-Learning Algorithm-Selection Models by Evolving Adversarial Instances.

Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches (2023)
Journal Article
Alissa, M., Sim, K., & Hart, E. (2023). Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches. Journal of Heuristics, 29(1), 1-38. https://doi.org/10.1007/s10732-022-09505-4

We propose a novel technique for algorithm-selection, applicable to optimisation domains in which there is implicit sequential information encapsulated in the data, e.g., in online bin-packing. Specifically we train two types of recurrent neural netw... Read More about Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches.

Algorithm selection using deep learning without feature extraction (2019)
Presentation / Conference Contribution
Alissa, M., Sim, K., & Hart, E. (2019, July). Algorithm selection using deep learning without feature extraction. Presented at Genetic and Evolutionary Computing Conference (GECCO) 2019, Prague, Czech Republic

We propose a novel technique for algorithm-selection which adopts a deep-learning approach, specifically a Recurrent-Neural Network with Long-Short-Term-Memory (RNN-LSTM). In contrast to the majority of work in algorithm-selection, the approach does... Read More about Algorithm selection using deep learning without feature extraction.

Use of machine learning techniques to model wind damage to forests (2018)
Journal Article
Hart, E., Sim, K., Kamimura, K., Meredieu, C., Guyon, D., & Gardiner, B. (2019). Use of machine learning techniques to model wind damage to forests. Agricultural and forest meteorology, 265, 16-29. https://doi.org/10.1016/j.agrformet.2018.10.022

This paper tested the ability of machine learning techniques, namely artificial neural networks and random forests, to predict the individual trees within a forest most at risk of damage in storms. Models based on these techniques were developed i... Read More about Use of machine learning techniques to model wind damage to forests.

A new rich vehicle routing problem model and benchmark resource (2018)
Presentation / Conference Contribution
Sim, K., Hart, E., Urquhart, N. B., & Pigden, T. (2015, September). A new rich vehicle routing problem model and benchmark resource. Presented at International Conference on Evolutionary and Deterministic Methods for Design, Optimization and Control with Applications to Industrial and Societal Problems, EUROGEN-2015, University of Strathclyde, Glasgow

We describe a new rich VRP model that captures many real-world constraints, following a recently proposed taxonomy that addresses both scenario and problem physical characteristics. The model is used to generate 4800 new instances of rich VRPs which... Read More about A new rich vehicle routing problem model and benchmark resource.

A hybrid method for feature construction and selection to improve wind-damage prediction in the forestry sector (2017)
Presentation / Conference Contribution
Hart, E., Sim, K., Gardiner, B., & Kamimura, K. (2017, July). A hybrid method for feature construction and selection to improve wind-damage prediction in the forestry sector. Presented at Proceedings of the Genetic and Evolutionary Computation Conference on - GECCO '17

Catastrophic damage to forests resulting from major storms has resulted in serious timber and financial losses within the sector across Europe in the recent past. Developing risk assessment methods is thus one of the keys to finding forest management... Read More about A hybrid method for feature construction and selection to improve wind-damage prediction in the forestry sector.

On Constructing Ensembles for Combinatorial Optimisation (2017)
Journal Article
Hart, E., & Sim, K. (2018). On Constructing Ensembles for Combinatorial Optimisation. Evolutionary Computation, 26(1), 67-87. https://doi.org/10.1162/evco_a_00203

Although the use of ensemble methods in machine-learning is ubiquitous due to their proven ability to outperform their constituent algorithms, ensembles of optimisation algorithms have received relatively little attention. Existing approaches lag beh... Read More about On Constructing Ensembles for Combinatorial Optimisation.

A hyper-heuristic ensemble method for static job-shop scheduling. (2016)
Journal Article
Hart, E., & Sim, K. (2016). A hyper-heuristic ensemble method for static job-shop scheduling. Evolutionary Computation, 24(4), 609-635. https://doi.org/10.1162/EVCO_a_00183

We describe a new hyper-heuristic method NELLI-GP for solving job-shop scheduling problems (JSSP) that evolves an ensemble of heuristics. The ensemble adopts a divide-and-conquer approach in which each heuristic solves a unique subset of the instance... Read More about A hyper-heuristic ensemble method for static job-shop scheduling..

A Novel Heuristic Generator for JSSP Using a Tree-Based Representation of Dispatching Rules (2015)
Presentation / Conference Contribution
Sim, K., & Hart, E. (2015, July). A Novel Heuristic Generator for JSSP Using a Tree-Based Representation of Dispatching Rules. Presented at Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference - GECCO Companion '15

A previously described hyper-heuristic framework named
NELLI is adapted for the classic Job Shop Scheduling Problem (JSSP) and used to find ensembles of reusable heuristics that cooperate to cover the heuristic search space. A new heuristic generato... Read More about A Novel Heuristic Generator for JSSP Using a Tree-Based Representation of Dispatching Rules.

A research agenda for metaheuristic standardization. (2015)
Presentation / Conference Contribution
Hart, E., & Sim, K. (2015, June). A research agenda for metaheuristic standardization. Paper presented at 11th Metaheuristics International Conference

We propose that the development of standardized, explicit, machine-readable descriptions of metaheuris- tics will greatly advance scientific progress in the field. In particular, we advocate a purely functional description of metaheuristics — separat... Read More about A research agenda for metaheuristic standardization..

A Lifelong Learning Hyper-heuristic Method for Bin Packing (2015)
Journal Article
Hart, E., Sim, K., & Paechter, B. (2015). A Lifelong Learning Hyper-heuristic Method for Bin Packing. Evolutionary Computation, 23(1), 37-67. https://doi.org/10.1162/EVCO_a_00121

We describe a novel Hyper-heuristic system which continuously learns over time to solve a combinatorial optimisation problem. The system continuously generates new heuristics and samples problems from its environment; representative problems and heur... Read More about A Lifelong Learning Hyper-heuristic Method for Bin Packing.

On the life-long learning capabilities of a NELLI*: a hyper-heuristic optimisation system. (2014)
Presentation / Conference Contribution
Hart, E., & Sim, K. (2014, September). On the life-long learning capabilities of a NELLI*: a hyper-heuristic optimisation system

Real-world applications of optimisation techniques place more importance on finding approaches that result in acceptable quality solutions in a short time-frame and can provide robust solutions, capable of being modified in response to changes in the... Read More about On the life-long learning capabilities of a NELLI*: a hyper-heuristic optimisation system..

A real-world employee scheduling and routing application. (2014)
Presentation / Conference Contribution
Hart, E., Sim, K., & Urquhart, N. B. (2014, July). A real-world employee scheduling and routing application. Presented at GECCO 2014

We describe a hyper-heuristic application developed for a client to find quick, acceptable solutions to Workforce Schedul- ing and Routing problems. An interactive fitness function controlled by the user enables five different objectives to be weight... Read More about A real-world employee scheduling and routing application..

An improved immune inspired hyper-heuristic for combinatorial optimisation problems. (2014)
Presentation / Conference Contribution
Sim, K., & Hart, E. (2014, July). An improved immune inspired hyper-heuristic for combinatorial optimisation problems

The meta-dynamics of an immune-inspired optimisation sys- tem NELLI are considered. NELLI has previously shown to exhibit good performance when applied to a large set of optimisation problems by sustaining a network of novel heuristics. We address th... Read More about An improved immune inspired hyper-heuristic for combinatorial optimisation problems..

Learning to solve bin packing problems with an immune inspired hyper-heuristic. (2013)
Presentation / Conference Contribution
Sim, K., Hart, E., & Paechter, B. (2013, September). Learning to solve bin packing problems with an immune inspired hyper-heuristic

Motivated by the natural immune system's ability to defend the body by generating and maintaining a repertoire of antibodies that collectively cover the potential pathogen space, we describe an artificial system that discovers and maintains a reperto... Read More about Learning to solve bin packing problems with an immune inspired hyper-heuristic..

Generating single and multiple cooperative heuristics for the one dimensional bin packing problem using a single node genetic programming island model. (2013)
Presentation / Conference Contribution
Sim, K., & Hart, E. (2013, July). Generating single and multiple cooperative heuristics for the one dimensional bin packing problem using a single node genetic programming island model. Presented at 15th annual conference on Genetic and evolutionary computation

Novel deterministic heuristics are generated using Single Node Genetic Programming for application to the One Dimensional Bin Packing Problem. First a single deterministic heuristic was evolved that minimised the total number of bins used when applie... Read More about Generating single and multiple cooperative heuristics for the one dimensional bin packing problem using a single node genetic programming island model..

A Hyper-Heuristic classifier for one dimensional bin packing problems: Improving classification accuracy by attribute evolution. (2012)
Presentation / Conference Contribution
Sim, K., Hart, E., & Paechter, B. (2012, September). A Hyper-Heuristic classifier for one dimensional bin packing problems: Improving classification accuracy by attribute evolution. Presented at International Conference on Parallel Problem Solving from Nature

A hyper-heuristic for the one dimensional bin packing problem is presented that uses an Evolutionary Algorithm (EA) to evolve a set of attributes that characterise a problem instance. The EA evolves divisions of variable quantity and dimension that r... Read More about A Hyper-Heuristic classifier for one dimensional bin packing problems: Improving classification accuracy by attribute evolution..

Outputs (17)