检索结果-内蒙古大学图书馆

IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering (ElConRus)

作者： Krymov, Roman A. Khamukhin, Anatoly, V RnD Ctr ELVEES Moscow Russia

ISBN: (纸本)9781665404761

Computer vision algorithms have complex set of parameters that significantly influence resulting quality. On the one hand, such a big amount of parameters that can be varied results in algorithm ability to fit for particular usage circumstances. On the other hand, fine-tuning of even several parameters may require a lot of human resources unless this task is automated. An approach to such automatization is provided considering a machine learning problem with several peculiarities. First, some parameters are restricted to have integer values. Second, multiple metrics describing algorithm quality are restricted to have integer values as well. Third, there are no analytical formulae describing dependences between parameters and metrics values. Mentioned features make gradient-descent-like methods widely used for neural networks training inapplicable for solving that optimization problem. As result, automatically tuned parameters provides better quality than manually selected ones.

关键词： machine learning hyperparameter optimization video analysis algorithm algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

Modular Differential Evolution 23

Modular Differential Evolution

引用

Genetic and Evolutionary Computation Conference (GECCO)

作者： Vermetten, Diederick Caraffini, Fabio Kononova, Anna V. Back, Thomas Leiden Inst Adv Comp Sci Leiden Netherlands Swansea Univ Swansea Wales

ISBN: (纸本)9798400701191

New contributions in the field of iterative optimisation heuristics are often made in an iterative manner. Novel algorithmic ideas are not proposed in isolation, but usually as extensions of a preexisting algorithm. Although these contributions are often compared to the base algorithm, it is challenging to make fair comparisons between larger sets of algorithm variants. This happens because even small changes in the experimental setup, parameter settings, or implementation details can cause results to become incomparable. Modular algorithms offer a way to overcome these challenges. By implementing the algorithmic modifications into a common framework, many algorithm variants can be compared, while ensuring that implementation details match in all versions. In this work, we propose a version of a modular framework for the popular Differential Evolution (DE) algorithm. We show that this modular approach not only aids in comparison but also allows for a much more detailed exploration of the space of possible DE variants. This is illustrated by showing that tuning the settings of modular DE vastly outperforms a set of commonly used DE versions which have been recreated in our framework. We then investigate these tuned algorithms in detail, highlighting the relation between modules and performance on particular problems.

关键词： Differential Evolution Benchmarking Modular algorithms algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

A Case Study on Automatic Parameter Optimization of a Mobile Robot Localization algorithm

A Case Study on Automatic Parameter Optimization of a Mobile...

引用

17th IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC)

作者： Lima, Oscar Ventura, Rodrigo Inst Super Tecn Inst Syst & Robot Ave Rovisco Pais 1 P-1049001 Lisbon Portugal

ISBN: (纸本)9781509062348

algorithms in robotics typically tend to expose several parameters for the user to configure. This allows both reusability and fine tuning of the algorithm to a particular setup, but at the expense of significant effort in tuning, since this task is typically done manually. However, recent results in parameter optimization have shown to be quite successful, namely in automatic configuration of boolean satisfiability problem solvers, contributing to a significant increase in scalability. In this paper we address the applicability of these methods to the area of mobile robot localization. In particular, we applied a sequential model-based optimization method to the automatic parameter tuning of the well-known Adaptive Monte Carlo Localization algorithm. Our results show a statistically significant improvement over the default algorithm values. We also contribute with an open source experimental setup, based on the popular Robot Operating System ROS, which can be easily adapted to other algorithms.

关键词： Model based optimization robotics black box optimization algorithm configuration automatic algorithm optimization multi objective parameter optimization automatic parameter tuning

来源：评论

学校读者我要写书评

暂无评论

Verification in the water phantom of the irradiation time calculation done by the algorithm used in intraoperative radiotherapy

引用

REPORTS OF PRACTICAL ONCOLOGY AND RADIOTHERAPY 2010年第5期15卷 132-137页

作者： Slyk, Maria Litoborski, Marcin Greater Poland Canc Ctr Dept Med Phys Poznan Poland

Aim: The investigation of the irradiation time calculation accuracy of the GGPB algorithm used for IORT. Background: Conventionally, breast conserving therapy consists of breast conserving surgery followed by postoperative whole breast irradiation and boost. The use of intraoperative radiotherapy (IORT) enables the boost to be delivered already during the surgery. In this case, the treatment dose for IORT can be calculated by use of General Gaussian Pencil Beam (GGPB) algorithm, which is implemented in TPS Eclipse. Materials and methods: PDDs and OFs for electron beams from Mobetron and all available applicators were measured in order to configure the GGPB algorithm. Afterwards, the irradiation times for the prescribed dose of 3Gy were calculated by means of it. The results of calculations were verified in the water phantom using the Marcus ionization chamber. Results: The results differed between energies. For 6MeV the irradiation times calculated by the GGPB algorithm were correct, for the energy of 9MeV they were too small and for the energy of 4MeV theywere too large for applicators with smaller diameters, while acceptable for the remaining ones. Conclusion: The GGPB algorithm can be used in intraoperative radiotherapy for energy and applicator sets for which no significant difference between the measured and the prescribed dose was obtained. For the rest of energy-applicator sets the configuration should be verified and possibly repeated. (C) 2010 Greater Poland Cancer Centre, Poland. Published by Elsevier Urban & Partner Sp. z.o.o. All rights reserved

关键词： GGPB algorithm algorithm configuration algorithm verification IORT

来源：评论

学校读者我要写书评

暂无评论

Reinforcement learning based adaptive metaheuristics 22

Reinforcement learning based adaptive metaheuristics

引用

Genetic and Evolutionary Computation Conference (GECCO)

作者： Tessari, Michele Iacca, Giovanni Univ Trento Trento Italy

ISBN: (纸本)9781450392686

Parameter adaptation, that is the capability to automatically adjust an algorithm's hyperparameters depending on the problem being faced, is one of the main trends in evolutionary computation applied to numerical optimization. While several handcrafted adaptation policies have been proposed over the years to address this problem, only few attempts have been done so far at apply machine learning to learn such policies. Here, we introduce a general-purpose framework for performing parameter adaptation in continuous-domain metaheuristics based on state-of-the-art reinforcement learning algorithms. We demonstrate the applicability of this framework on two algorithms, namely Covariance Matrix Adaptation Evolution Strategies (CMA-ES) and Differential Evolution (DE), for which we learn, respectively, adaptation policies for the step-size (for CMA-ES), and the scale factor and crossover rate (for DE). We train these policies on a set of 46 benchmark functions at different dimensionalities, with various inputs to the policies, in two settings: one policy per function, and one global policy for all functions. Compared, respectively, to the Cumulative Step-size Adaptation (CSA) policy and to two well-known adaptive DE variants (iDE and jDE), our policies are able to produce competitive results in the majority of cases, especially in the case of DE.

关键词： Evolutionary algorithms Reinforcement Learning Adaptation algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

Adaptive Differential Evolution based on Exploration and Exploitation Control

Adaptive Differential Evolution based on Exploration and Exp...

引用

IEEE Congress on Evolutionary Computation (IEEE CEC)

作者： Bai, Hao Huang, Changwu Yao, Xin INSA Rouen Normandy Lab Mech Normandy LMN F-76000 Rouen France Southern Univ Sci & Technol Dept Comp Sci & Engn Guangdong Prov Key Lab Brain Inspired Intelligent Shenzhen 518055 Peoples R China

ISBN: (纸本)9781728183923

Search operator design and parameter tuning are essential parts of algorithm design. However, they often involve trial-and-error and are very time-consuming. A new differential evolution (DE) algorithm with adaptive exploration and exploitation control (AEEC-DE) is proposed in this work to tackle this challenge. The proposed method improves the performance of DE by automatically selecting trial vector generation strategies (both mutation and crossover operators) and dynamically generating the associated control parameter values. A probability-based exploration and exploitation measurement is introduced to estimate whether the state of each newly generated individual is in exploration or exploitation. The state of historical individuals is used to assess the exploration and exploitation capabilities of different generation strategies and parameter values. Then, the strategies and parameters of DE are adapted following the common belief that evolutionary algorithms (EAs) should start with exploration and then gradually change into exploitation. The performance of AEEC-DE is evaluated through experimental studies on a set of test problems and compared with several state-of-the-art adaptive DE variants.

关键词： algorithm configuration Differential Evolution Parameter Control Exploration and Exploitation

来源：评论

学校读者我要写书评

暂无评论

Configuring the Perturbation Operations of an Iterated Local Search algorithm for Cross-domain Search: A Probabilistic Learning Approach

Configuring the Perturbation Operations of an Iterated Local...

引用

IEEE Congress on Evolutionary Computation (IEEE CEC)

作者： Adubi, Stephen A. Oladipupo, Olufunke O. Olugbara, Oludayo O. Covenant Univ Dept Comp & Informat Sci Ota Nigeria Durban Univ Technol Dept Informat Technol ICT & Soc Res Grp Durban South Africa

ISBN: (纸本)9781728183923

Hyper-heuristics are general-purpose heuristic search methodologies for solving combinatorial optimization problems (COPs). Research findings have revealed that hyper-heuristics still suffer generalization issues as different strategies vary in performance from an instance of a COP to another. In this paper, an approach based on Iterated Local Search (ILS) is proposed to raise the level of generality of hyper-heuristics on the problem domains of the HyFlex framework. The proposed approach utilizes a probabilistic learning technique to automatically configure the behavior of the ILS algorithm during the perturbation stage of the optimization process. In the proposed method, the mutation and ruin-recreate heuristics are treated as distinct entities and the learning layer automatically determines the level of utilization of these heuristic categories depending on the problem domain being solved. The concept of double shaking is also presented where a solution can be perturbed twice before the intensification phase. Experimental results reveal the level of generality achieved by the proposed method as it recorded a minimum formula one score of 30.0 on each tested problem domain. Direct comparison with a state-of-the-art ILS-based approach also establishes the significance of the learning layer added to the perturbation stage of the ILS metaheuristic. Finally, analysis of the perturbation behavior of the hyper-heuristic leads to an interesting conclusion concerning the type of low-level heuristics that are highly beneficial and non-beneficial to a given problem domain.

关键词： iterated local search selection hyper-heuristics probabilistic learning algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

Configuring Mixed-Integer Programming Solvers for Large-Scale Instances

引用

Operations Research Forum 2024年第2期5卷 48页

作者： Kemminer, Robin Lange, Jannick Kempkes, Jens Peter Tierney, Kevin Weiß, Dimitri OPTANO GmbH Technologiepark 18 NRW Paderborn 33100 Germany Decision and Operation Technologies Bielefeld University Universitätsstraße 25 NRW Bielefeld 33615 Germany

algorithm configuration techniques automatically search for parameters of solvers and algorithms that provide minimal runtime or maximal solution quality on specified instance sets. Mixed-integer programming (MIP) solvers pose a particular challenge for algorithm configurators due to the difficulty of finding optimal, or even feasible, solutions on the large-scale problems commonly found in practice. We introduce the OPTANO algorithm Tuner (OAT) to find configurations for MIP solvers and other optimization algorithms. We present and evaluate several critical components of OAT for solving MIPs in particular and show that OAT can find configurations that significantly improve the solution time of MIPs on two different datasets. © The Author(s) 2024.

关键词： algorithm configuration Large-scale problem instances Mixed-integer programming

来源：评论

学校读者我要写书评

暂无评论

New Aspects of Beyond Worst-Case Analysis

New Aspects of Beyond Worst-Case Analysis

引用

作者： Colin White Carnegie Mellon University

学位级别：博士

Traditionally, the theory of algorithms has focused on the analysis of worst-case instances. While this has led to beautiful results and a thorough understanding of a wide variety of problems, there are many problems for which worst-case analysis is not useful for empirical or real-world instances. A rapidly developing line of research, the so-called beyond worst-case analysis of algorithms (BWCA), considers the design and analysis of problems using more realistic models or using natural structural properties. The goals of BWCA are to design realistic models and new theoretical algorithms which perform well in practice, as well as to explain why preexisting algorithms that do well in practice perform better than worst-case analysis suggests. In other words, the goal of BWCA is to bridge the gap between theory and practice. In this thesis, we continue the line of work of BWCA for clustering and distributed learning by making contributions in several areas. Specifically, we focus on three main problems and models. • Clustering is one problem that has benefited greatly from BWCA. The perturba- tion resilience assumption proposed by Bilu and Linial (2011) states that the op- timal clustering does not change under small perturbations to the input distances. We design efficient algorithms that output optimal or near-optimal clusterings for the canonical k-center objective (in both symmetric and asymmetric forms) under perturbation resilience. Our results are tight up to the level of perturbation resilience assumed. Next, we give robust algorithms with guarantees even if only part of the data satisfies the perturbation resilience assumption. • Clustering has many varied applications in practice, and an algorithm may have drastically different performance between two applications. To reconcile this fact, we consider a data-driven approach, in which a clustering application is represented by a distribution over problem instances, and the goal is to find the best algorithm fo

关键词： Beyond Worst-Case Analysis Clustering k-means Perturbation Resilience Distributed Computation algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

Automated configuration and Usage of Strategy Portfolios for Mixed-Motive Bargaining 22

Automated Configuration and Usage of Strategy Portfolios for...

引用

Proceedings of the 21st International Conference on Autonomous Agents and Multiagent Systems

作者： Bram M. Renting Holger H. Hoos Catholijn M. Jonker Leiden University & Delft University of Technology Leiden Netherlands RWTH Aachen & Leiden University Aachen Germany Delft University of Technology & Leiden University Delft Netherlands

ISBN: (纸本)9781450392136

Bargaining can be used to resolve mixed-motive games in multi-agent systems. Although there is an abundance of negotiation strategies implemented in automated negotiating agents, most agents are based on single fixed strategies, while it is acknowledged that there is no single best-performing strategy for all negotiation *** this paper, we focus on bargaining settings where opponents are repeatedly encountered, but the bargaining problems change. We introduce a novel method that automatically creates and deploys a portfolio of complementary negotiation strategies using a training set and optimise pay-off in never-before-seen bargaining settings through per-setting strategy selection. Our method relies on the following contributions. We introduce a feature representation that captures characteristics for both the opponent and the bargaining problem. We model the behaviour of an opponent during a negotiation based on its actions, which is indicative of its negotiation strategy, in order to be more effective in future *** combination of feature-based methods generalises to new negotiation settings, as in practice, over time, it selects effective counter strategies in future encounters. Our approach is tested in an ANAC-like tournament, and we show that we are capable of winning such a tournament with a (5.6%) increase in pay-off compared to the runner-up agent.

关键词： mixed-motive games bargaining algorithm selection algorithm configuration

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：