检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

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

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

4 篇 会议

馆藏范围

4 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

2 篇 工学
- 2 篇 计算机科学与技术...
2 篇 管理学
- 2 篇 管理科学与工程(可...

主题

4 篇 computer aided/a...
3 篇 evolutionary rob...
3 篇 artificial life
2 篇 genetic algorith...
2 篇 evolving virtual...
1 篇 multi-objective ...
1 篇 deep learning
1 篇 morphogenesis
1 篇 parallel optimiz...

机构

1 篇 ucsd ece dept la...
1 篇 charles universi...
1 篇 scuola super san...
1 篇 univ bristol dep...
1 篇 computer center ...

作者

1 篇 peter krčah
1 篇 koushanfar farin...
1 篇 javidi tara
1 篇 calisti marcello
1 篇 laschi cecilia
1 篇 javaheripi mojan
1 篇 peter krah
1 篇 corucci francesc...
1 篇 hauser helmut
1 篇 samragh mohammad

语言

4 篇 英文

检索条件"主题词=Computer aided/automated design"

共 4 条记录，以下是1-10 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

GeneCAI: Genetic Evolution for Acquiring Compact AI

GeneCAI: Genetic Evolution for Acquiring Compact AI

引用

Genetic and Evolutionary Computation Conference (GECCO)

作者： Javaheripi, Mojan Samragh, Mohammad Javidi, Tara Koushanfar, Farinaz UCSD ECE Dept La Jolla CA 92093 USA

ISBN: (纸本)9781450371285

In the contemporary big data realm, Deep Neural Networks (DNNs) are evolving towards more complex architectures to achieve higher inference accuracy. Model compression techniques can be lever-aged to efficiently deploy these compute-intensive architectures on resource-limited mobile devices. Such methods comprise various hyperparameters that require per-layer customization to ensure high accuracy. Choosing the hyperparameters is cumbersome as the pertinent search space grows exponentially with model layers. This paper introduces GeneCAI, a novel optimization method that automatically learns how to tune per-layer compression hyperparameters. We devise a bijective translation scheme that encodes compressed DNNs to the genotype space. Each genotype's optimality is measured using a multi-objective score based on the accuracy and number of floating-point operations. We develop customized genetic operations to iteratively evolve the non-dominated solutions towards the optimal Pareto front, thus, capturing the optimal trade-off between model accuracy and complexity. GeneCAI optimization method is highly scalable and can achieve a near-linear performance boost on distributed multi-GPU platforms. Our extensive evaluations demonstrate that GeneCAI outperforms existing rule-based and reinforcement learning methods in DNN compression by finding models that lie on a better accuracy/complexity Pareto curve.

关键词： Deep Learning Genetic Algorithms Multi-objective Optimization computer aided/automated design Parallel Optimization

来源：评论

学校读者我要写书评

暂无评论

Novelty-Based Evolutionary design of Morphing Underwater Robots 15

Novelty-Based Evolutionary Design of Morphing Underwater Rob...

引用

17th Genetic and Evolutionary Computation Conference (GECCO)

作者： Corucci, Francesco Calisti, Marcello Hauser, Helmut Laschi, Cecilia Scuola Super Sant Anna BioRobot Inst Pisa Italy Univ Bristol Dept Engn Math Bristol BS8 1TH Avon England

ISBN: (纸本)9781450334723

Recent developments in robotics demonstrated that bioinspiration and embodiement are powerful tools to achieve robust behavior in presence of little control. In this context morphological design is usually performed by humans, following a set of heuristic principles: in general this can be limiting, both from an engineering and an artificial life perspectives. In this work we thus suggest a different approach, leveraging evolutionary techniques. The case study is the one of improving the locomotion capabilities of an existing bioinspired robot. First, we explore the behavior space of the robot to discover a number of qualitatively different morphology-enabled behaviors, from whose analysis design indications are gained. The suitability of novelty search a recent open-ended evolutionary algorithm for this intended purpose is demonstrated. Second, we show how it is possible to condense such behaviors into a reconfigurable robot capable of online morphological adaptation (morphosis, morphing). Examples of successful morphing are demonstrated, in which changing just one morphological parameter entails a dramatic change in the behavior: this is promising for a future robot design. The approach here adopted represents a novel computed-aided, bioinspired, design paradigm, merging human and artificial creativity. This may result in interesting implications also for artificial life, having the potential to contribute in exploring underwater locomotion "as-it-could-be".

关键词： computer aided/automated design morphogenesis artificial life evolutionary robotics genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Towards efficient evolution of morphology and control 08

Towards efficient evolution of morphology and control

引用

Proceedings of the 10th annual conference on Genetic and evolutionary computation

作者： Peter Krah Computer Center Charles University Prague Czech Rep

ISBN: (纸本)9781605581309

We propose a novel algorithm for the evolution of body and control of three-dimensional, physically simulated virtual creatures controlled by artificial neural networks. The proposed algorithm is inspired by NeuroEvolution of Augmenting Topologies (NEAT) which efficiently evolves artificial neural networks. All three main components of NEAT algorithm (protecting evolutionary innovation through speciation, effective crossover of neural networks with different topologies and incremental growth from minimal structure) are applied to the evolution of both morphology and control system of the virtual creatures. Large-scale experiments have shown that the proposed algorithm evolves creatures using significantly less fitness evaluations than a standard genetic algorithm on all four tested fitness functions. Positive contribution of each component of the proposed algorithm has been confirmed with a series of supplementary ablation experiments.

关键词： evolutionary robotics evolving virtual creatures artificial life computer aided/automated design

来源：评论

学校读者我要写书评

暂无评论

Evolving virtual creatures revisited 07

Evolving virtual creatures revisited

引用

Proceedings of the 9th annual conference on Genetic and evolutionary computation

作者： Peter Krčah Charles University Prague Czech Republic

ISBN: (纸本)9781595936974

Thirteen years have passed since Karl Sims published his work onevolving virtual creatures. Since then,several novel approaches toneural network evolution and genetic algorithms have been *** aim of our work is to apply recent results in these areas tothe virtual creatures proposed by Karl Sims, leading to creaturescapable of solving more complex tasks. This paper presents oursuccess in reaching the first milestone -a new and completeimplementation of the original virtual creatures. All morphologicaland control properties of the original creatures were *** of physics are simulated using ODE library. Distributedcomputation is used for CPU-intensive tasks, such as *** have shown that our system is capable ofevolving both morphology and control of the creatures resulting ina variety of non-trivial swimming and walking strategies.

关键词： computer aided/automated design evolving virtual creatures evolutionary robotics artificial life

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共1页 << < 1 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

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：