检索结果-内蒙古大学图书馆

Backdoors in Neural models of Source Code 26

Backdoors in Neural Models of Source Code

26th international Conference on Pattern Recognition / 8th international workshop on Image Mining - theory and Applications (IMTA)

作者： Ramakrishnan, Goutham Albarghouthi, Aws Hlth Scale Corp San Jose CA 95113 USA Univ Wisconsin Madison Madison WI USA

ISBN: (数字)9781665490627

ISBN: (纸本)9781665490627

Deep neural networks are vulnerable to a range of adversaries. A particularly pernicious class of vulnerabilities are backdoors, where model predictions diverge in the presence of subtle triggers in inputs. An attacker can implant a backdoor by poisoning the training data to yield a desired target prediction on triggered inputs. We study backdoors in the context of deep-learning for source code. (1) We define a range of backdoor classes for source-code tasks and install backdoors using dataset poisoning. (2) We adapt and improve recent algorithms from robust statistics for our setting, showing that backdoors leave a spectral signature in the learned representation of source code, thus enabling detection of poisoned data. (3) We conduct a thorough evaluation on different architectures and languages, showing the ease of injecting backdoors and our ability to eliminate them.

关键词： Deep learning Codes Source coding Neural networks Training data Implants Predictive models

来源：评论

学校读者我要写书评

暂无评论

database Schema Evolution and Meta-Modeling : 9th international workshop on foundations of models and languages for data and objects, Fomlado/Demm 2000, Dagstuhl Castle, Germany, September 18-21, 2000 : Selected Papers

引用

2016年

作者： herman balsters

database Schema Evolution and Meta-Modeling : 9th international workshop on foundations of models and languages for data and objects, Fomlado/Demm 2000, Dagstuhl Castle, Germany, September 18-21, 2000 : Selected Papers by international workshop on foundations of models; languages for data; objects (9th : 2000 Dagstuhl, Wadern, Germany); Balsters, Herman, 1952-; Brock, Bert De; Conrad, S. (Stefan); published by Berlin ; New York : Springer

关键词： castle, dagstuhl database database management object-oriented databases

来源：评论

学校读者我要写书评

暂无评论

data MODELING FOR MUSEUM COLLECTIONS 8

DATA MODELING FOR MUSEUM COLLECTIONS

引用

8th international workshop on 3D Virtual Reconstruction and Visualization of Complex Architectures (3D-ARCH)

作者： Lo Turco, M. Calvano, M. Giovannini, E. C. Politecn Torino Dept Architecture & Design Turin Italy

the relationship between cultural heritage, digital technologies and visual models involves an increasingly wide area of research, oriented towards the renewal of archives and museums for the preservation and promotion of culture. Recent research activities are the result of the progressive strengthening of digital technologies and the needs of a new generation of "digital" users, which requires museums to update their means of communication using Semantic Web languages and technologies shaped by a social conceptualization of a graph-based representation of information. the growth of several digitized heritage collections increases the necessity of proper methodologies to develop a structured system able to access to these collections and the large amount of data, metadata and paradata related to the digitized objects in a structured and organized way, defining a set of collection information models (CIM), that considers not only the digitizing process but also the data collection process, layered by an Upper Ontology level structure, based on CIDOC-CRM.

关键词： 3D Modelling data Modeling Museum Collections Archival Research Paradata Documentation

来源：评论

学校读者我要写书评

暂无评论

the logic of query languages for data streams 11

The logic of query languages for data streams

引用

Proceedings of the 4th international workshop on Logic in databases

作者： Zaniolo, Carlo Computer Science Department University of California at Los Angeles Los Angeles CA United States

ISBN: (纸本)9781450306096

data Stream Management Systems (DSMS) represent a vibrant research area that is rich in technical challenges, which many projects have approached by extending database query languages and models for continuous queries on data streams [1, 3, 4, 9, 5]. these database-inspired approaches have delivered remarkable systems and applications, but have yet to produce solid conceptual foundations for DSMS data models and query languages|particularly if we compare with the extraordinary ones of relational databases. A cornerstone of the success of relational databases was the tight coupling between their data model and their logic-based query languages. In this paper, we show that a similar approach can succeed for data streams and propose a tight-coupled design for DSMS data models and query languages. To express more naturally the behavior of a data stream and attain more powerful on-line queries, we abandon the set-of-tuples model of relational databases, and instead use sequences of tuples ordered by their time-stamps as our data stream model. this approach allows us to overcome the blocking problem that severely impairs the expressive power of data stream query languages. As elucidated in [1]: A blocking query operator is one that cannot produce the first tuple of the output until it has seen the entire input. Previous work had characterized blocking query operators by their nonmonotonic behavior [7, 6, 8]. In this paper, we instead use the closed-world assumption [11, 10] to characterize blocking/ nonblocking behaviors with respect to the incompleteness/ completeness of the streaming database. From this, we infer simple syntactic conditions that make datalog rules immune from blocking. A significant and surprising new result is that the use of negated goals in the bodies of rules does not imply a blocking behavior: in fact, many very useful nonblocking queries can be expressed using negation. the ip side of this exciting result is that additional conditions must then be

关键词： Query languages

来源：评论

学校读者我要写书评

暂无评论

8th international workshop on foundations of models and languages for data and objects, FoMLaDO 1999

8th International Workshop on Foundations of Models and Lang...

引用

8th international workshop on foundations of models and languages for data and objects, FMLDO 1999

ISBN: (纸本)9783540672012

the proceedings contain 13 papers. the special focus in this conference is on foundations of models and languages for data and objects. the topics include: Federated transaction management with snapshot isolation;serializability, concurrency control, and replication control;logical update queries as open nested transactions;inheritance in a deductive object database language with updates;specifying distributed and dynamically evolving information systems using an extended co-nets approach;specifying active rules for database maintenance;nested transactions with integrity constraints;declarative specifications of complex transactions;a support for increasing performance and multi-tier applications;execution guarantees in electronic commerce payments;transactions in mobile electronic commerce;transactions and electronic commerce and overview and discussion topics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

9th international workshop on foundations of models and languages for data and objects, workshop on database Schema Evolution and Meta-Modeling, FoMLaDO/DEMM 2000

引用

9th international workshop on foundations of models and languages for data and objects, FoMLaDO/DEMM 2000

ISBN: (纸本)3540422722

the proceedings contain 13 papers. the special focus in this conference is on database Schema Evolution and Meta-Modeling. the topics include: Consistency management in runtime evolving concurrent information systems;adaptive specifications of technical information systems;evolving the software of a schema evolution system;schema evolution and versioning;a logical and computational characterisation;temporal branching as a conflict management technique;QFD matrix for incremental construction of a warehouse via data marts;change propagation in an axiomatic model of schema evolution for objectbase management systems;conceptual description of adaptive information systems;extending the object query language for transparent metadata access;a metamodeling approach to evolution and defining metrics for conceptual schema evolution.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Expressiveness of structured document query languages based on attribute grammars

引用

JOURNAL OF thE ACM 2002年第1期49卷 56-100页

作者： Neven, F Van den Bussche, J Limburgs Univ Centrum Dept WNI B-3590 Diepenbeek Belgium

Structured document databases can be naturally viewed as derivation trees of a context-free grammar. Under this view, the classical formalism of attribute grammars becomes a formalism for structured document query languages, From this perspective, we studs the expressive power of BAGs: Boolean-valued attribute grammars with propositional logic formulas as semantic rules, and RAGs: relation-valued attribute grammars with first-order logic formulas as semantic rules. BAGs can express only unary queries;RAGs can express queries of any arity. We first show that the (unary) queries expressible by BAGs are precisely those definable in monadic second-order logic, We then show that the queries expressible by RAGs are precisely those definable by first-order inductions of linear depth, or, equivalently, those computable in linear time on a parallel machine with polynomially many processors. Further, we show that RAGs that only use synthesized attributes are strictly weaker than RAGs that use both synthesized and inherited attributes. We show that RAGs are more expressive than monadic second-order logic for queries of any arity. Finally, we discuss relational attribute grammars in the context of BAGs and RAGs, We show that in the case of BAGs this does not increase the expressive power, while different semantics for relational RAGs capture the complexity classes NP, coNP and UP boolean AND coUP.

关键词： languages theory attribute grammars automata complexity logic structured documents

来源：评论

学校读者我要写书评

暂无评论

Qfd matrix for incremental construction of a warehouse via data marts 9th

引用

9th international workshop on foundations of models and languages for data and objects, FoMLaDO/DEMM 2000

作者： McFadyen, Ron Chan, Fung-Yee University of Winnipeg Winnipeg Canada

ISBN: (纸本)3540422722

In this paper we consider the construction of a dimensional data warehouse. the warehouse is built beginning with the first data mart and proceeding in an iterative manner constructing one mart at a time. In this way the warehouse is seen to evolve over time. this evolutionary process is necessary due to the complexity of data stores, relationships, transformations, and the processing involved. In this paper we consider the problem of identifying the next data mart to construct and present a tool based on Quality Function Deployment for use in the planning stages. © Springer-Verlag Berlin Heidelberg 2001.

关键词： data warehouses

来源：评论

学校读者我要写书评

暂无评论

database schema evolution and meta-modeling

引用

丛书名： Lecture notes in computer science,

2001年

作者： Herman Balsters Bert de Brock Stefan Conrad (eds.).

来源：评论

学校读者我要写书评

暂无评论

Evolving objects: Conceptual description of adaptive information systems 9th

引用

9th international workshop on foundations of models and languages for data and objects, FoMLaDO/DEMM 2000

作者： Saake, Gunter Türker, Can Conrad, Stefan Otto-von-Guericke-Universität Magdeburg Institut für Technische und Betriebliche Informationssysteme D-39016 MagdeburgPostfach 4120 Germany Zurich Institute of Information Systems ETH Zentrum ZurichCH-8092 Switzerland Ludwig-Maximilians-Universität München Institut für Informatik Oettingenstr 67 MünchenD-80538 Germany

ISBN: (纸本)3540422722

Today, information systems are essential parts of large organizations. Since such kinds of systems have a very long life-span, they have to be adapted to new changing requirements occurring during their lifetime. Evolution must be regarded not only at the object state level, but also at the object behavior level. Especially, the explicit handling of (behavior) evolution on the conceptual level is necessary. For that, we introduce the notion of evolving objects as basic building blocks of information systems. the behavior of such an object is divided into a rigid and an evolving part. the rigid behavior is ideally stable for the whole life-span of the object;the evolving behavior can be changed dynamically at runtime. In this paper, we present an extended specification framework for modeling evolving objects. Particularly, this framework provides the basis to explicitly specify behavior evolution. © Springer-Verlag Berlin Heidelberg 2001.

关键词： Information systems

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：