Kevin Sim
YOU?
Author Swipe
View article: Class Incremental Learning for Algorithm Selection
Class Incremental Learning for Algorithm Selection Open
Algorithm selection is commonly used to predict the best solver from a portfolio per per-instance. In many real scenarios, instances arrive in a stream: new instances become available over time, while the number of class labels can also gr…
View article: Beyond the Hype: Benchmarking LLM-Evolved Heuristics for Bin Packing
Beyond the Hype: Benchmarking LLM-Evolved Heuristics for Bin Packing Open
Coupling Large Language Models (LLMs) with Evolutionary Algorithms has recently shown significant promise as a technique to design new heuristics that outperform existing methods, particularly in the field of combinatorial optimisation. An…
View article: Evaluating the Robustness of Deep-Learning Algorithm-Selection Models by Evolving Adversarial Instances
Evaluating the Robustness of Deep-Learning Algorithm-Selection Models by Evolving Adversarial Instances Open
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 ev…
View article: To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features
To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features Open
Dynamic algorithm selection aims to exploit the complementarity of multiple optimization algorithms by switching between them during the search. While these kinds of dynamic algorithms have been shown to have potential to outperform their …
View article: Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches
Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches Open
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 …
View article: To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features - Dataset
To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features - Dataset Open
This repository contains the reproduction steps and intermediate artifacts corresponding to the paper 'To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features'. During the submission …
View article: To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features - Dataset
To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features - Dataset Open
This repository contains the reproduction steps and intermediate artifacts corresponding to the paper 'To Switch or not to Switch: Predicting the Benefit of Switching between Algorithms based on Trajectory Features'. During the submission …
View article: Minimising line segments in linear diagrams is NP-hard
Minimising line segments in linear diagrams is NP-hard Open
Linear diagrams have been shown to be an effective method of representing set-based data. Moreover, a number of guidelines have been proven to improve the efficacy of linear diagrams. One of these guidelines is to minimise the number of li…
View article: Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches
Automated Algorithm Selection: from Feature-Based to Feature-Free Approaches Open
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 …
View article: Evolutionary Approaches to Improving the Layouts of Instance-Spaces
Evolutionary Approaches to Improving the Layouts of Instance-Spaces Open
We propose two new methods for evolving the layout of an instance-space. Specifically we design three different fitness metrics that seek to: (i) reward layouts which place instances won by the same solver close in the space; (ii) reward l…
View article: A Neural Approach to Generation of Constructive Heuristics
A Neural Approach to Generation of Constructive Heuristics Open
Both algorithm-selection methods and hyper-heuristic methods rely on a pool of complementary heuristics. Improving the pool with new heuristics can improve performance, however, designing new heuristics can be challenging. Methods such as …
View article: A deep learning approach to predicting solutions in streaming optimisation domains
A deep learning approach to predicting solutions in streaming optimisation domains Open
In the field of combinatorial optimisation, per-instance algorithm selection still remains a challenging problem, particularly with respect to streaming problems such as packing or scheduling. Typical approaches involve training a model to…
View article: Algorithm selection using deep learning without feature extraction
Algorithm selection using deep learning without feature extraction Open
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 app…
View article: Use of machine learning techniques to model wind damage to forests
Use of machine learning techniques to model wind damage to forests Open
View article: A hybrid method for feature construction and selection to improve wind-damage prediction in the forestry sector
A hybrid method for feature construction and selection to improve wind-damage prediction in the forestry sector Open
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…
View article: On Constructing Ensembles for Combinatorial Optimisation
On Constructing Ensembles for Combinatorial Optimisation Open
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 approac…
View article: A Hyper-Heuristic Ensemble Method for Static Job-Shop Scheduling
A Hyper-Heuristic Ensemble Method for Static Job-Shop Scheduling Open
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 t…
View article: A Novel Heuristic Generator for JSSP Using a Tree-Based Representation of Dispatching Rules
A Novel Heuristic Generator for JSSP Using a Tree-Based Representation of Dispatching Rules Open
A previously described hyper-heuristic framework namedNELLI 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…
View article: Roll Project Bin Packing Benchmark Problems
Roll Project Bin Packing Benchmark Problems Open
This document describes two sets of Benchmark Problem Instances for the One Dimensional Bin Packing Problem. The problem instances are supplied as compressed (zipped) SQLITE database files
View article: Roll project job shop scheduling benchmark problems
Roll project job shop scheduling benchmark problems Open
This document describes two sets of benchmark problem instances for the job shop scheduling problem. Each set of instances is supplied as a compressed (zipped) archive containing a single CSV file for each problem instance using the format…
View article: Roll project rich vehicle routing benchmark problems
Roll project rich vehicle routing benchmark problems Open
This document describes a large set of Benchmark Problem Instances for the Rich Vehicle Routing Problem. All files are supplied as a single compressed (zipped) archive containing the instances, in XML format, an Object-Oriented Model suppl…
View article: Genetic programming : 18th European conference, EuroGP 2015, Copenhagen, Denmark, April 8-10, 2015, proceedings
Genetic programming : 18th European conference, EuroGP 2015, Copenhagen, Denmark, April 8-10, 2015, proceedings Open