检索结果-内蒙古大学图书馆

Performance Comparison of Software Reliability Estimation algorithms

COMPUTER 2024年第4期57卷 26-36页

作者： Yano, Hiromu Dohi, Tadashi Okamura, Hiroyuki Hiroshima Univ Adv Sci & Engn Hiroshima 7390046 Japan

Specific optimization algorithms have been developed for the purpose of automated software reliability assessment tools. In this article, we propose the Monte Carlo expectation-maximization algorithm as another optimization algorithm, and carry out the performance comparison of the software reliability estimation algorithms through comprehensive numerical experiments.

关键词： Monte Carlo methods Software algorithms Estimation Software reliability expectation-maximization algorithms Optimization

来源：评论

学校读者我要写书评

暂无评论

Asymptotic Errors for Teacher-Student Convex Generalized Linear Models (Or: How to Prove Kabashima's Replica Formula)

引用

IEEE TRANSACTIONS ON INFORMATION THEORY 2023年第3期69卷 1824-1852页

作者： Gerbelot, Cedric Abbara, Alia Krzakala, Florent Univ PSL Sorbonne Univ Lab Phys Ecole normale Super CNRS F-75005 Ens Paris France NYU Courant Inst Math Sci New York NY 10012 USA Lab Phys Ecole normalesu perieure EPFL F-75005 Ens Paris France Lab Computat Biol & Theoret Biophys EPFL CH-1015 Lausanne Switzerland IdePh Lab EPFL CH-1015 Lausanne Switzerland

There has been a recent surge of interest in the study of asymptotic reconstruction performance in various cases of generalized linear estimation problems in the teacher-student setting, especially for the case of i.i.d standard normal matrices. Here, we go beyond these matrices, and prove an analytical formula for the reconstruction performance of convex generalized linear models with rotationally-invariant data matrices with arbitrary bounded spectrum, rigorously confirming, under suitable assumptions, a conjecture originally derived using the replica method from statistical physics. The proof is achieved by leveraging on message passing algorithms and the statistical properties of their iterates, allowing to characterize the asymptotic empirical distribution of the estimator. For sufficiently strongly convex problems, we show that the two-layer vector approximate message passing algorithm (2-MLVAMP) converges, where the convergence analysis is done by checking the stability of an equivalent dynamical system, which gives the result for such problems. We then show that, under a concentration assumption, an analytical continuation may be carried out to extend the result to convex (non-strongly) problems. We illustrate our claim with numerical examples on mainstream learning methods such as sparse logistic regression and linear support vector classifiers, showing excellent agreement between moderate size simulation and the asymptotic prediction.

关键词： Heuristic algorithms Approximation algorithms Estimation Convergence Signal processing algorithms Mathematical models Machine learning Parametric statistics estimation error message passing expectation-maximization algorithms optimization convergence convex functions linear matrix inequalities

来源：评论

学校读者我要写书评

暂无评论

Inferring structure of cortical neuronal networks from activity data: A statistical physics approach

引用

PNAS NEXUS 2025年第1期4卷 pgae565页

作者： Po, Ho Fai Houben, Akke Mats Haeb, Anna-Christina Jenkins, David Rhys Hill, Eric J. Parri, H. Rheinallt Soriano, Jordi Saad, David UNIV ASTON DEPT PHARM BIRMINGHAM B4 7ET England Univ Barcelona Dept Fis Mat Condensada E-08028 Barcelona Spain Univ Barcelona Inst Complex Syst UBICS Barcelona 08028 Spain Aston Univ Coll Hlth & Life Sci Birmingham B4 7ET England Aston Univ Aston Inst Membrane Excellence Birmingham B4 7ET England Loughborough Univ Dept Chem Loughborough LE11 3TU Leics England

Understanding the relation between cortical neuronal network structure and neuronal activity is a fundamental unresolved question in neuroscience, with implications to our understanding of the mechanism by which neuronal networks evolve over time, spontaneously or under stimulation. It requires a method for inferring the structure and composition of a network from neuronal activities. Tracking the evolution of networks and their changing functionality will provide invaluable insight into the occurrence of plasticity and the underlying learning process. We devise a probabilistic method for inferring the effective network structure by integrating techniques from Bayesian statistics, statistical physics, and principled machine learning. The method and resulting algorithm allow one to infer the effective network structure, identify the excitatory and inhibitory type of its constituents, and predict neuronal spiking activity by employing the inferred structure. We validate the method and algorithm's performance using synthetic data, spontaneous activity of an in silico emulator, and realistic in vitro neuronal networks of modular and homogeneous connectivity, demonstrating excellent structure inference and activity prediction. We also show that our method outperforms commonly used existing methods for inferring neuronal network structure. Inferring the evolving effective structure of neuronal networks will provide new insight into the learning process due to stimulation in general and will facilitate the development of neuron-based circuits with computing capabilities.

关键词： biological neuronal networks inference neuronal-type classification kinetic Ising model generalized maximum likelihood expectation-maximization algorithms

来源：评论

学校读者我要写书评

暂无评论

Covariance-Free Sparse Bayesian Learning

引用

IEEE TRANSACTIONS ON SIGNAL PROCESSING 2022年 70卷 3818-3831页

作者： Lin, Alexander Song, Andrew H. Bilgic, Berkin Ba, Demba Harvard Univ Sch Engn & Appl Sci Cambridge MA 02138 USA MIT Elect Engn & Comp Sci Dept Cambridge MA 02138 USA MIT Harvard MIT Hlth Sci & Technol Cambridge MA 02138 USA Athinoula A Martinos Ctr Biomed Imaging Charlestown MA 02129 USA Harvard Med Sch Dept Radiol Boston MA 02115 USA

Sparse Bayesian learning (SBL) is a powerful framework for tackling the sparse coding problem while also providing uncertainty quantification. The most popular inference algorithms for SBL exhibit prohibitively large computational costs for high-dimensional problems due to the need to maintain a large covariance matrix. To resolve this issue, we introduce a new method for accelerating SBL inference - named covariance-free expectation maximization (CoFEM) - that avoids explicit computation of the covariance matrix. CoFEM solves multiple linear systems to obtain unbiased estimates of the posterior statistics needed by SBL. This is accomplished by exploiting innovations from numerical linear algebra such as preconditioned conjugate gradient and a little-known diagonal estimation rule. For a large class of compressed sensing matrices, we provide theoretical justifications for why our method scales well in high-dimensional settings. Through simulations, we show that CoFEM can be up to thousands of times faster than existing baselines without sacrificing coding accuracy. Through applications to calcium imaging deconvolution and multi-contrast MRI reconstruction, we show that CoFEM enables SBL to tractably tackle high-dimensional sparse coding problems of practical interest.

关键词： Signal processing algorithms Inference algorithms Bayes methods Encoding Covariance matrices Linear systems Computational modeling Sparse Bayesian learning (SBL) compressed sensing expectation-maximization algorithms computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Time Series Impact Through Topic Modeling

引用

IEEE ACCESS 2022年 10卷 97327-97347页

作者： Cendrero, Julian Gonzalo, Julio Galletero, Marcos Zapata, Ivar MrHouston Tech Solut Dept Artificial Intelligence Madrid 28002 Spain Univ Nacl Educ Distancia UNED NLP & IR Res Grp Madrid 28040 Spain Univ Complutense Madrid UCM Dept Mat Phys Madrid 28040 Spain

A time-series of numerical data and a sequence of time-ordered documents are often correlated. This paper aims at modeling the impact that the underlying themes discussed in the text data have on the time series. To do so, we introduce an original topic model, Time Series Impact Through Topic Modeling (TSITM), that includes contextual data by coupling Latent Dirichlet Allocation (LDA) with linear regression, using an elastic net prior to set to zero the impact of uncorrelated topics. The resulting topics act as explanatory variables for the regression of the numerical time series, which allows us to understand the time series movements based on the events described on the text data. We have tested our model on two datasets: first, we used political news to explain the US president's disapproval ratings;then, we considered a corpus of economic news to explain the financial returns of 4 different multinational corporations. Our experiments show that an appropriate selection of hyperparameters (via repeated random subsampling validation and Bayesian optimization) leads to significant correlations: both an intrinsic baseline and state of the art methods were significantly outperformed by TSITM in MSE, MAE and out-of-sample R-2, according to our hypothesis tests. We believe that this framework can be useful in the context of reputational risk management.

关键词： expectation-maximization algorithms natural language processing regression analysis text mining text analysis time series analysis

来源：评论

学校读者我要写书评

暂无评论

Research of Russian Conversion Algorithm based on Hidden Markov Model

Research of Russian Conversion Algorithm based on Hidden Mar...

引用

Image Processing, Electronics and Computers (IPEC), Asia-Pacific Conference on

作者： Yang Wang School of International Education Heilongjiang Institute of Technology JiXi China

ISBN: (数字)9798350374407

ISBN: (纸本)9798350374414

Phonetic conversion technology is pivotal in developing resources for Russian speech information processing. This paper aims to enhance the Russian phoneme set based on SAMPA, ensuring accurate reflection of stress position and vowel reduction in Russian words. Utilizing the improved phoneme set, this paper successfully constructed a Russian pronunciation dictionary with 20,000 words. A data-driven Russian phonetic conversion algorithm leveraging the Hidden Markov Model (HMM) for alignment, modeling, and decoding is proposed. Initially, the expectation-maximization algorithm aligns Russian phonetics in a “many-to-many” manner. Subsequently, alignment results are trained using a joint N-gram model and transformed into a HMM pronunciation model. The HMM decoding algorithm ultimately predicts word pronunciation accurately. Cross-validation experiments yielded satisfactory results, and they show that the proposed can be easily implemented in the computer assisted system and provide enhanced accuracy and convenience to the Russian conversion process.

关键词： Accuracy Dictionaries Transducers Computational modeling Hidden Markov models Phonetics Prediction algorithms Decoding expectation-maximization algorithms Stress

来源：评论

学校读者我要写书评

暂无评论

Research on Language Recognition algorithms Based on Segment-Level Features and Automatic Identification

Research on Language Recognition Algorithms Based on Segment...

引用

IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA)

作者： Le Zu Yong Kang Jie Gao Shanshan He School of General Education Wannan Medical College Wuhu China

ISBN: (数字)9798350377255

ISBN: (纸本)9798350377262

Language recognition technology plays a crucial role in automated speech processing within multilingual environments, particularly under the globalized context where the increasing linguistic diversity poses higher demands on language recognition systems. Due to their robustness and efficiency, segment-level features and automatic identification algorithms have gradually become focal points of research. This paper explores a variety of language signal feature parameters, including RASTA-PLP parameters, MFCC parameters, RASTA-MFCC parameters, and CMS-MFCC parameters, as well as extraction methods for sliding difference cepstrum parameters and their applications in language recognition. Gaussian Mixture Models, along with their improved expectation-maximization algorithms and greedy expectation-maximization algorithms, are thoroughly analyzed to enhance the accuracy and robustness of language recognition. Preliminary experiments verify the performance of different feature parameters and algorithms on actual datasets, demonstrating that RASTA and CMS techniques significantly enhance the MFCC parameters.

关键词： Accuracy Training data Speech recognition Feature extraction Robustness Multilingual Mel frequency cepstral coefficient expectation-maximization algorithms Optimization Gaussian mixture model

来源：评论

学校读者我要写书评

暂无评论

Statistical Model for Battery Recharging Time in RIS-assisted WPT Systems

Statistical Model for Battery Recharging Time in RIS-assiste...

引用

Communications and Networking (MECOM), IEEE Middle East Conference on

作者： Rawan Derbas Shimaa Naser Sami Muhaidat 6G Research Center Khalifa University Abu Dhabi United Arab Emirates Department of Systems and Computer Engineering Carleton University Ottawa ON Canada

ISBN: (数字)9798350376715

ISBN: (纸本)9798350376722

In this paper, we propose an analytical framework to statistically analyze the battery recharging time (BRT) in reconfigurable intelligent surfaces (RISs)-assisted wireless power transfer (WPT) systems. Specifically, we derive novel expressions for the probability density function (PDF), and cumulative distribution function (CDF) of the BRT of the radio frequency energy harvesting (RFEH) wireless nodes. Considering both the direct link and the links provided by the RIS, the received power at the RFEH nodes is modeled as a gamma mixture model (GMM), where the parameters of GMM are calculated using the expectation-maximization algorithm. Utilizing the derived expressions, we provide a comprehensive analysis for the statistical characterization of the BRT and investigate how the system and battery parameters affect the charging time performance. Finally, simulation results are provided to verify the theoretical results.

关键词： Wireless communication Statistical analysis Simulation Wireless power transfer Mixture models Reconfigurable intelligent surfaces Probability density function Batteries Energy harvesting expectation-maximization algorithms

来源：评论

学校读者我要写书评

暂无评论

Direction of Arrival Estimation on a Sphere

Direction of Arrival Estimation on a Sphere

引用

International Workshop on Acoustic Signal Enhancement (IWAENC)

作者： Alexis Favrot Christof Faller Illusonic GmbH Switzerland

ISBN: (数字)9798350361858

ISBN: (纸本)9798350361865

The direction of arrival (DOA) of a sound source emitting in the three dimensional (3D) space can be defined by both its azimuth and elevation angles. Estimating these two angles, using a multiple microphone array, enables to build a two dimensional (2D) histogram to track sources, and such depict the current audio scene. The observation data collected in the histogram are analyzed and modeled, providing multiple, varying DOA values. An analytic function, which fulfills periodicity and continuity over a sphere, is proposed to parametrize the 2D histogram as a statistical model in order to run an iterative algorithm, such as the expectation-maximization (EM) one, providing eventually the corresponding DOA values.

关键词： Histograms Analytical models Solid modeling Direction-of-arrival estimation Three-dimensional displays Two-dimensional displays Estimation Microphone arrays expectation-maximization algorithms Distribution functions

来源：评论

学校读者我要写书评

暂无评论

Approximation of the expectation-maximization algorithm for Gaussian mixture models on big data

Approximation of the expectation-maximization algorithm for ...

引用

IEEE International Conference on Big Data

作者： Mateusz Przyborowski Dominik Ślęzak Institute of Informatics University of Warsaw QED Software Warsaw Poland

ISBN: (纸本)9781665480468

Gaussian mixture models are a very useful tool for modeling data distribution. While estimating parameters using the expectation-maximization algorithm, this approach does not scale well with big datasets, especially if it is necessary to prepare many models for the proper selection of metaparameters. In this article we present an approximation of the expectation-maximization algorithm obtained by merging crucial subsets of the dataset, that differ slightly in their effect on the expectation-maximization loss function, into information granules. Furthermore, application examples comparing new method with the classical approach are shown.

关键词： Computational modeling Merging Big Data Approximation algorithms Data models expectation-maximization algorithms Gaussian mixture model

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：