检索结果-内蒙古大学图书馆

On w-mixtures: Finite convex combinations of prescribed component distributions

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Nielsen, Frank Nock, Richard Sony Computer Science Laboratories Inc Japan Data61 The Australian National University The University of Sydney Australia

We consider the space of w-mixtures which is defined as the set of finite statistical mixtures sharing the same prescribed component distributions closed under convex combinations. The information geometry induced by the Bregman generator set to the Shannon negentropy on this space yields a dually flat space called the mixture family manifold. We show how the Kullback-Leibler (KL) divergence can be recovered from the corresponding Bregman divergence for the negentropy generator: That is, the KL divergence between any two w-mixtures amounts to a Bregman Divergence (BD) induced by the Shannon negentropy generator. Thus the KL divergence between two Gaussian Mixture Models (GMMs) sharing the same Gaussian components is equivalent to a Bregman divergence. This KL-BD equivalence on a mixture family manifold implies that we can perform optimal KL-averaging aggregation of w-mixtures without information loss. More generally, we prove that the statistical skew Jensen-Shannon divergence between w-mixtures is equivalent to a skew Jensen divergence between their corresponding parameters. Finally, we state several properties, divergence identities, and divergence inequalities relating to w-mixtures. Copyright © 2017, The Authors. All rights reserved.

关键词： Machine learning

Fast Hybrid Transactional Index with Scalable Phantom Avoidance

学校读者我要写书评

暂无评论

Fast Hybrid Transactional Index with Scalable Phantom Avoida...

International Symposium on Computing and Networking Workshops (CANDARW)

作者： Yutaro Bessho Hideyuki Kawashima Computer and Data Science Laboratories NTT Tokyo Japan Faculty of Environment and Information Studies Keio University Fujisawa Japan

ISBN: (数字)9798331505349

ISBN: (纸本)9798331505356

Griffin is a hybrid transactional index structure that combines a hash table and B+-tree. The key benefit of Griffin is that it processes single-key operations fast using the hash table. For serializable transactions, Griffin avoids phantoms with precision locking, a technique that protects updates and predicate reads with a lock. Although Griffin performs well under infrequent updates, under update-heavy workloads, the lists of locks kept by precision locking significantly grow in size, causing severe performance degradation. We identify this is because locks are released by only one thread called sync manager, which is also responsible for the heavy task of synchronizing the hash table and B+-tree. As a result, sync manager often fails to release locks as fast as they are acquired, causing the lock lists to grow. We propose a new design for Griffin, Griffin-PPL (Parallel Precision Locking), which prevents the lock lists from growing. Griffin-PPL removes sync manager, the bottlenecking thread, and parallelizes the work of releasing locks and synchronization among all transaction worker threads. To allow for parallelization, we adopt a decentralized data structure for the lock lists. Our evaluation in a microbenchmark shows that under an update-heavy workload (50% inserts, 50% scans), Griffin-PPL exhibits a throughput 6.5x the conventional Griffin architecture, at 64 threads.

关键词： Degradation databases Conferences Phantoms Parallel processing data structures Throughput Synchronization Indexes

Towards Log-based Execution Status Estimation Using Graph Neural Networks

学校读者我要写书评

暂无评论

Towards Log-based Execution Status Estimation Using Graph Ne...

Asia-Pacific Conference on Software Engineering

作者： Shimon Sumita Hiroyuki Nakagawa Shinobu Saito Tatsuhiro Tsuchiya Osaka University Osaka Japan Okayama University Okayama Japan NTT Computer and Data Science Laboratories Tokyo Japan

ISBN: (数字)9798331534011

ISBN: (纸本)9798331534028

This study addresses software bloat, a prevalent issue in modern software development, causing excessive size and complexity due to feature additions and unnecessary functions. Such bloat leads to decreased efficiency, performance degradation, and increased vulnerability. To combat this issue, the concept of software 3R (reduce, reuse, recycle) is proposed; however, accurately reproducing the internal state of black-box software for 3R requires both source code and execution log data, posing practical challenges. In this paper, we conduct a software execution status estimation using limited execution log. Graph Neural Networks (GNNs) are employed for analysis, offering effective processing of graph data. The task is framed as link prediction and node classification, comparing traditional deep learning methods with GNNs using Apache OFBiz ERP software logs. Preliminary results validate GNN applicability.

关键词： Degradation Deep learning Source coding Estimation Closed box Graph neural networks Recycling Complexity theory Software engineering Software development management

Griffin: Fast Transactional database Index with Hash and B+-Tree

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Nakazono, Sho Bessho, Yutaro Kawashima, Hideyuki Nakamori, Tatsuhiro Computer and Data Science Laboratories NTT Tokyo Japan Faculty of Environment and Information Studies Keio University Kanagawa Japan

Index access is one of the dominant performance factors in transactional database systems. Many systems use a B+tree or one of its variants to handle point and range operations. This access pattern has room for performance improvement. Firstly, point operations can potentially be processed in O(1) with a hash table. Secondly, to ensure serializability of transactions, range operations incur overhead from phantom avoidance techniques that involve additional processing or synchronization, such as an extra traversal of the B+-tree. To address these issues, we propose a hybrid index architecture, Griffin. For point operations, Griffin has a hash table that provides access paths in O(1) time, along with a B+-tree. For phantom avoidance, Griffin employs a precision locking method, which does not involve additional traversal of the B+-tree. Despite its hybrid architecture, Griffin transparently provides linearizable operations and an interface of a single database index. We built a Griffin index combining a hash table and BwTree. Compared to a baseline index that is composed of a BwTree only, it achieves up to 3.1x higher throughput in a point operation dominant workload, and up to 5.4x higher throughput in a range operation dominant workload. © 2024, CC BY.

关键词： database systems

ON THE GEOMETRY OF MIXTURES OF PRESCRIBED DISTRIBUTIONS

学校读者我要写书评

暂无评论

ON THE GEOMETRY OF MIXTURES OF PRESCRIBED DISTRIBUTIONS

IEEE International Conference on Acoustics, Speech and Signal Processing

作者： Frank Nielsen Richard Nock école Polytechnique France Sony Computer Science Laboratories Japan Data 61 Australia The Australian National & Sydney Universities

ISBN: (纸本)9781538646595

We consider the space of w-mixtures that are finite statistical mixtures sharing the same prescribed component distributions, like Gaussian mixture models sharing the same components. The information geometry induced by the Kullback-Leibler (KL) divergence yields a dually flat space where the KL divergence between two w-mixtures amounts to a Bregman divergence for the negative Shannon entropy generator, called the Shannon information. Furthermore, we prove that the skew Jensen-Shannon statistical divergence between w-mixtures amount to skew Jensen divergences on their parameters and state several divergence inequalities between wmixtures and their closures.

关键词： GEOMETRY MIXTURES PRESCRIBED DISTRIBUTIONS geometry mixtures Divergence divergence Divergence divergence Shannon information Information geometry surgical closure Skew

Griffin: Fast Transactional database Index with Hash and B+-Tree

学校读者我要写书评

暂无评论

Griffin: Fast Transactional Database Index with Hash and B+-...

IEEE International Conference on e-science and Grid Computing

作者： Sho Nakazono Yutaro Bessho Hideyuki Kawashima Tatsuhiro Nakamori Computer and Data Science Laboratories NTT Tokyo Japan Faculty of Environment and Information Studies Keio University Kanagawa Japan

ISBN: (数字)9798350365610

ISBN: (纸本)9798350365627

关键词： Phantoms Throughput data structures Indexes Synchronization Periodic structures

Type-Based Verification of Connectivity Constraints in Lattice Surgery

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wakizaka, Ryo Suzuki, Yasunari Igarashi, Atsushi Graduate School of Informatics Kyoto University Japan NTT Computer and Data Science Laboratories Musashino180-8585 Japan

Fault-tolerant quantum computation using lattice surgery can be abstracted as operations on graphs, wherein each logical qubit corresponds to a vertex of the graph, and multi-qubit measurements are accomplished by connecting the vertices with paths between them. Operations attempting to connect vertices without a valid path will result in abnormal termination. As the permissible paths may evolve during execution, it is necessary to statically verify that the execution of a quantum program can be completed. This paper introduces a type-based method to statically verify that well-typed programs can be executed without encountering halts induced by surgery operations. Alongside, we present QLS, a first-order quantum programming language to formalize the execution model of surgery operations. Furthermore, we provide a type checking algorithm by reducing the type checking problem to the offline dynamic connectivity problem. © 2024, CC BY.

关键词： Qubits

Clustering in Hilbert simplex geometry

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Nielsen, Frank Sun, Ke Sony Computer Science Laboratories Inc. Japan Data61 Australia The Australian National University Australia

Clustering categorical distributions in the finite-dimensional probability simplex is a fundamental task met in many applications dealing with normalized histograms. Traditionally, the differential-geometric structures of the probability simplex have been used either by (i) setting the Riemannian metric tensor to the Fisher information matrix of the categorical distributions, or (ii) defining the dualistic information-geometric structure induced by a smooth dissimilarity measure, the Kullback-Leibler divergence. In this work, we introduce for clustering tasks a novel computationally-friendly framework for modeling geometrically the probability simplex: The Hilbert simplex geometry. In the Hilbert simplex geometry, the distance is the non-separable Hilbert's metric distance which satisfies the property of information monotonicity with distance level set functions described by polytope boundaries. We show that both the Aitchison and Hilbert simplex distances are norm distances on normalized logarithmic representations with respect to the 2 and variation norms, respectively. We discuss the pros and cons of those different statistical modelings, and benchmark experimentally these different kind of geometries for center-based k-means and k-center clustering. Furthermore, since a canonical Hilbert distance can be defined on any bounded convex subset of the Euclidean space, we also consider Hilbert's geometry of the elliptope of correlation matrices and study its clustering performances compared to Fröbenius and log-det divergences. Copyright © 2017, The Authors. All rights reserved.

关键词： Geometry

Runtime Semantic Interoperability for Gathering Ontology-based Network Context

学校读者我要写书评

暂无评论

Runtime Semantic Interoperability for Gathering Ontology-bas...

IEEE Symposium on Network Operations and Management

作者： J. Keeney D. Lewis D. O'Sullivan A. Roelens V. Wade A. Boran R. Richardson Knowledge & Data Engineering Group Department of Computer Science Trinity College Dublin Dublin Ireland Bell Laboratories Ireland

The trends for pushing more operational intelligence towards network elements to achieve more context-aware and self-managing behavior often requires elements to gather network knowledge without necessarily binding explicitly to all of the potential sources of that knowledge. Though event-based publish-subscribe models allow efficient distribution of knowledge where the event types are known globally, dynamic service chains, ad hoc networks and pervasive computing application all introduce a more fluid and heterogeneous range of context knowledge. This requires some runtime translation of knowledge between sources and sinks of network context. This paper builds on existing mapping techniques that use ontological forms of existing management information models to examine the extent to which these can be employed for runtime semantic interoperability for network knowledge. It presents results in developing a management knowledge delivery framework based on existing models and platforms, but which offers a more decentralized knowledge exchange mechanism

关键词： Runtime Ontologies Knowledge management Intelligent networks Publish-subscribe Context modeling Fluid dynamics Ad hoc networks Pervasive computing Context-aware services