检索结果-内蒙古大学图书馆

Naturally Interpretable Control Policies via graph-based genetic programming 27th

Naturally Interpretable Control Policies via Graph-Based Gen...

27th European Conference on genetic programming (EuroGP) Held as Part of EvoStar Conference27th European Conference on genetic programming (EuroGP)

作者： Nadizar, Giorgia Medvet, Eric Wilson, Dennis G. Univ Trieste Dept Math & Geosci Trieste Italy Univ Trieste Dept Engn & Architecture Trieste Italy Univ Toulouse ISAE SUPAERO Toulouse France

ISBN: (纸本)9783031569562;9783031569579

In most high-risk applications, interpretability is crucial for ensuring system safety and trust. However, existing research often relies on hard-to-understand, highly parameterized models, such as neural networks. In this paper, we focus on the problem of policy search in continuous observations and actions spaces. We leverage two graph-based genetic programming (GP) techniques-Cartesian genetic programming (CGP) and Linear genetic programming (LGP)-to develop effective yet interpretable control policies. Our experimental evaluation on eight continuous robotic control benchmarks shows competitive results compared to state-of-the-art Reinforcement Learning (RL) algorithms. Moreover, we find that graph-based GP tends towards small, interpretable graphs even when competitive with RL. By examining these graphs, we are able to explain the discovered policies, paving the way for trustworthy AI in the domain of continuous control.

关键词： graph-based genetic programming Cartesian genetic programming Linear genetic programming Interpretable Policy Continuous Control

来源：评论

学校读者我要写书评

暂无评论

Searching for a Diversity of Interpretable graph Control Policies

Searching for a Diversity of Interpretable Graph Control Pol...

引用

genetic and Evolutionary Computation Conference (GECCO)

作者： Nadizar, Giorgia Medvet, Eric Wilson, Dennis G. Univ Trieste MIGe Trieste Italy Univ Trieste DIA Trieste Italy Univ Toulouse ISAE Supaero Toulouse France

ISBN: (纸本)9798400704949

graph-based genetic programming (GGP) can create interpretable control policies in graph form, but faces challenges such as local optima and solution fragility, which undermine its efficacy. Quality-Diversity (QD) has been effective in addressing similar issues, traditionally in Artificial Neural Network (ANN) optimization. In this paper, we introduce a general graph Quality-Diversity (G-QD) framework to enhance the performance of GGP with QD optimization, obtaining a variety of interpretable, effective, and resilient policies. Using Cartesian genetic programming (CGP) as the GGP technique and MAP-Elites (ME) as the QD algorithm, we leverage a combination of behavior and graph structural descriptors. Experimenting on two navigation and two locomotion continuous control tasks, our framework yields an array of effective yet behaviorally and structurally diverse policies, surpassing the performance of a standard genetic Algorithm (GA). The resulting solution set also increases interpretability, allowing for insight into the control tasks. Additionally, our experiments demonstrate the robustness of the solutions to faults such as sensor damage.

关键词： graph-based genetic programming Cartesian genetic programming Quality-Diversity MAP-Elites Interpretable Policy Continuous Control

来源：评论

学校读者我要写书评

暂无评论

Horizontal gene transfer for recombining graphs

引用

genetic programming AND EVOLVABLE MACHINES 2020年第3期21卷 321-347页

作者： Atkinson, Timothy Plump, Detlef Stepney, Susan Univ York Dept Comp Sci York N Yorkshire England

We introduce a form of neutral horizontal gene transfer (HGT) to evolving graphs by graph programming (EGGP). We introduce the mu x lambda evolutionary algorithm (EA), where mu parents each produce lambda children who compete only with their parents. HGT events then copy the entire active component of one surviving parent into the inactive component of another parent, exchanging genetic information without reproduction. Experimental results from symbolic regression problems show that the introduction of the mu x lambda EA and HGT events improve the performance of EGGP. Comparisons with genetic programming and Cartesian genetic programming strongly favour our proposed approach. We also investigate the effect of using HGT events in neuroevolution tasks. We again find that the introduction of HGT improves the performance of EGGP, demonstrating that HGT is an effective cross-domain mechanism for recombining graphs.

关键词： graph-based genetic programming Neuroevolution Horizontal gene transfer

来源：评论

学校读者我要写书评

暂无评论

Dynamical genetic programming in XCSF

引用

EVOLUTIONARY COMPUTATION 2013年第3期21卷 361-387页

作者： Preen, Richard J. Bull, Larry Univ W England Dept Comp Sci & Creat Technol Bristol BS16 1QY Avon England

A number of representation schemes have been presented for use within learning classifier systems, ranging from binary encodings to artificial neural networks. This paper presents results from an investigation into using a temporally dynamic symbolic representation within the XCSF learning classifier system. In particular, dynamical arithmetic networks are used to represent the traditional condition-action production system rules to solve continuous-valued reinforcement learning problems and to perform symbolic regression, finding competitive performance with traditional genetic programming on a number of composite polynomial tasks. In addition, the network outputs are later repeatedly sampled at varying temporal intervals to perform multistep-ahead predictions of a financial time series.

关键词： graph-based genetic programming learning classifier systems multistep-ahead prediction reinforcement learning self-adaptation symbolic regression XCSF

来源：评论

学校读者我要写书评

暂无评论

Evolutionary algorithm considering program size: efficient program evolution using grape 08

Evolutionary algorithm considering program size: efficient p...

引用

Proceedings of the 10th annual conference companion on genetic and evolutionary computation

作者： Shinichi Shirakawa Tomoharu Nagao Yokohama National University Yokohama Kanagawa Japan

ISBN: (纸本)9781605581316

Today, a lot of Automatic programming techniques have been proposed and applied various fields. graph Structured Program Evolution (GRAPE) is one of the recent Automatic programming techniques. GRAPE succeeds in generating the complex programs automatically. In this paper, a new generation alternation model for GRAPE, called Evolutionary Algorithm Considering Program Size (EACP), is proposed. EACP maintains the diversity of program size in the population by using particular fitness assignment and generation alternation. We apply EACP to three test problems, factorial, exponentiation and sorting a list. And we show the effectiveness of EACP and confirm evolution of maintaining the diversity of program size.

关键词： automatic programming generation alternation model graph-based genetic programming genetic algorithm genetic programming evolutionary algorithm

来源：评论

学校读者我要写书评

暂无评论

graph Structured Program Evolution 07

Graph Structured Program Evolution

引用

Annual Conference of genetic and Evolutionary Computation Conference

作者： Shirakawa, Shinichi Ogino, Shintaro Nagao, Tomoharu Yokohama Natl Univ Grad Sch Environm & Informat Sci Hodogaya Ku 79-7 Tokiwadai Yokohama Kanagawa 2408501 Japan Yokohama Natl Univ Venture Business Lab Hodogaya Ku Yokohama Kanagawa 2408501 Japan

ISBN: (纸本)9781595936974

In recent Years a lot of Automatic programming techniques have developed. A typical example of Automatic programming is genetic programming (GP), and various extensions and representations for GP have been proposed so far. However, it seems that more improvements are necessary to obtain complex programs automatically. lit this paper we proposed a new method called graph Structured Program Evolution (GRAPE). The representation of GRAPE is graph structure, therefore it can represent complex programs (e.g. branches and loops) using its graph structure. Each program is constructed as an arbitrary directed graph of nodes and data set. The GRAPE program handles multiple data types rising the data set for each type, and the genotype of GRAPE is the form of a linear string of integers. We apply GRAPE to four test problems, factorial, Fibonacci sequence, exponentiation and reversing a list, and demonstrate that the optimum solution in each problem is obtained by the GRAPE system.

关键词： Automatic programming genetic programming graph-based genetic programming genetic Algorithm factorial Fibonacci sequence exponentiation reversing a list

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：