检索结果-内蒙古大学图书馆

The Journal of Machine Learning Research 2022年第1期23卷 14124-14160页

作者： Adam Block Zeyu Jia Yury Polyanskiy Alexander Rakhlin Department of Mathematics Massachusetts Institute of Technology Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology Department of Brain & Cognitive Sciences Statistics and Data Science Center Massachusetts Institute of Technology

It has long been thought that high-dimensional data encountered in many practical machine learning tasks have low-dimensional structure, i.e., the manifold hypothesis holds. A natural question, thus, is to estimate the intrinsic dimension of a given population distribution from a finite sample. We introduce a new estimator of the intrinsic dimension and provide finite sample, non-asymptotic guarantees. We then apply our techniques to get new sample complexity bounds for Generative Adversarial Networks (GANs) depending only on the intrinsic dimension of the data.

关键词： manifold hypothesis dimension estimation manifold learning intrinsic dimension Hölder GANs

来源：评论

学校读者我要写书评

暂无评论

Regret Minimization and Statistical Inference in Online Decision Making with High-dimensional Covariates

arXiv

引用

arXiv 2024年

作者： Duan, Congyuan Ma, Wanteng Jiang, Jiashuo Xia, Dong Department of Mathematics Hong Kong University of Science and Technology Hong Kong Department of Statistics and Data Science University of Pennsylvania United States Department of Industrial Engineering & Decision Analytics Hong Kong University of Science and Technology Hong Kong

This paper investigates regret minimization, statistical inference, and their interplay in high-dimensional online decision-making based on the sparse linear context bandit model. We integrate the Ε-greedy bandit algorithm for decision-making with a hard thresholding algorithm for estimating sparse bandit parameters and introduce an inference framework based on a debiasing method using inverse propensity weighting. Under a margin condition, our method achieves either O(T1/2) regret or classical O(T1/2)-consistent inference, indicating an unavoidable trade-off between exploration and exploitation. If a diverse covariate condition holds, we demonstrate that a pure-greedy bandit algorithm—i.e., exploration-free—combined with a debiased estimator based on average weighting can simultaneously achieve optimal O(log T) regret and O(T1/2)-consistent inference. We also show that a simple sample mean estimator can provide valid inference for the optimal policy’s value. Numerical simulations and experiments on Warfarin dosing data validate the effectiveness of our methods. © 2024, CC BY.

关键词： Inverse problems

来源：评论

学校读者我要写书评

暂无评论

Learning Causal Relations from Subsampled Time Series with Two Time-Slices 41

Learning Causal Relations from Subsampled Time Series with T...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Wu, Anpeng Li, Haoxuan Kuang, Kun Zhang, Keli Wu, Fei Department of Computer Science and Technology Zhejiang University Hangzhou China Center for Data Science Peking University Beijing China Huawei Noah's Ark Lab Huawei Shenzhen China Shanghai Institute for Advanced Study Zhejiang University Shanghai China Shanghai AI Laboratory Shanghai China

This paper studies the causal relations from subsampled time series, in which measurements are sparse and sampled at a coarser timescale than the causal timescale of the underlying system. In such data, because there are numerous missing time-slices (i.e., cross-sections at each time point) between two consecutive measurements, conventional causal discovery methods designed for standard time series data would produce significant errors. To learn causal relations from subsampled time series, a typical solution is to conduct different interventions and then make a comparison. However, full interventions are often expensive, unethical, or even infeasible, particularly in fields such as health and social science. In this paper, we first explore how readily available two-time-slices data can replace intervention data to improve causal ordering, and propose a novel Descendant Hierarchical Topology algorithm with Conditional Independence Test (DHT-CIT) to learn causal relations from subsampled time series using only two time-slices. Specifically, we develop a conditional independence criterion that can be applied iteratively to test each node from time series and identify all of its descendant nodes. Empirical results on both synthetic and real-world datasets demonstrate the superiority of our DHT-CIT algorithm. Copyright 2024 by the author(s)

关键词： Time series analysis

来源：评论

学校读者我要写书评

暂无评论

Transductive - Long Short-Term Memory Network for the Fake News Detection 3

Transductive - Long Short-Term Memory Network for the Fake N...

引用

3rd IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics, ICDCECE 2024

作者： Pillai, Sanjaikanth E Vadakkethil Somanathan Avacharmal, Rajiv Reddy, R. Archana Pareek, Piyush Kumar Zanke, Pankaj School Of Electrical Engineering And Computer Science University Of North Dakota Grand Forks United States Center For Data Science And Artificial Intelligence New York Life Insurance Company Farmington HillsMI United States School Of Computer Science And Engineering Sr University Telangana Warangal India Nitte Meenakshi Institute Of Technology Department Of Aiml And Ipr Cell Bengaluru India Sapiens International Corporation United States

ISBN: (纸本)9798350318609

Recently, fake news and rumors are distributing majorly and rapidly in all over world. That circumstance causes the production and propagation of incorrect news articles. As well as mis data and fake news are maximized through user without correct verification. In this manuscript, proposed a Transductive - Long Short-Term Memory (T-LSTM) method for detecting fake news. The Buzzfeed and ISOT datasets are used in this manuscript for fake news detection and it is pre-processed by tokenization, removal of stop word and stemming. Then, the reduction of feature is performed by using Probability Principal Component Analysis (PPCA) and it is classified by using T-LSTM method. The proposed T-LSTM method attained accuracy 96.51 %, precision 94.64 %, recall 92.33 % and f1-score 93.57% on Buzzfeed dataset. The proposed T-LSTM method attained accuracy 98.43%, precision 97.76%, recall 97.02% and f1-score 97.35% on ISOT dataset. The proposed T-LSTM method performed well than other existing methods like LSTM -Levy Flight (LSTM-LF) and OptNet. © 2024 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Iot Based Smart Assistant Cane For The Visually Impaired

Iot Based Smart Assistant Cane For The Visually Impaired

引用

2024 International Conference on Communication, Computing and Internet of Things, IC3IoT 2024

作者： Harini, A. Buvana, C. Harshini, M. Keerthana, S. Prabhu, D. Umamaheswaran, S.K. Sri Sairam Engineering College Department Of Electronics And Communication Engineering Tamil Nadu Chennai India Sri Sairam Engineering College Department Of Information Technology Tamil Nadu Chennai India Sri Sairam Engineering College Department Of Artificial Intelligence And Data Science Tamil Nadu Chennai India Sri Sairam Engineering College Department Of Computer Science And Business Systems Tamil Nadu Chennai India Sri Sairam Engineering College Department Of Mathematics Tamil Nadu Chennai India

ISBN: (纸本)9798350352689

This paper presents the design and implementation of a Smart Cane for the visually impaired individuals, incorporating several sensors and features to improve their mobility and safety. The device includes an ultrasonic sensor for obstacle detection, an MPU-6050 device for detecting falls and tilts, a GPS module for navigation, an RTC module for medicine reminders, and a fitness tracking module for monitoring health parameters. Additionally, a Bluetooth headset with a microphone is provided to communicate with the voice assistant in noisy areas. The Smart Cane triggers a buzzer for help and sends a notification to the caretaker in case of a fall or crash. The device is powered by a LIPO battery and is designed to improve the quality of life of visually impaired individuals by providing them with enhanced mobility and safety features. © 2024 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Impacts of Innovation School System in Korea: A Latent Space Item Response Model with Neyman-Scott Point Process

arXiv

引用

arXiv 2023年

作者： Yi, Seorim Kim, Minkyu Park, Jaewoo Jeon, Minjeong Jin, Ick Hoon Department of Statistics and Data Science Yonsei University Korea Republic of Department of Statistics and Data Science University of Texas United States Department of Applied Statistics Yonsei University Korea Republic of School of Education and Information Studies University of California Los Angeles United States

South Korea’s educational system has faced criticism for its lack of focus on critical thinking and creativity, resulting in high levels of stress and anxiety among students. As part of the government’s effort to improve the educational system, the innovation school system was introduced in 2009, which aims to develop students’ creativity as well as their non-cognitive skills. To better understand the differences between innovation and regular school systems in South Korea, we propose a novel method that combines the latent space item response model (LSIRM) with the Neyman-Scott (NS) point process model. Our method accounts for the heterogeneity of items and students, captures relationships between respondents and items, and identifies item and student clusters that can provide a comprehensive understanding of students’ behaviors/perceptions on non-cognitive outcomes. Our analysis reveals that students in the innovation school system show a higher sense of citizenship, while those in the regular school system tend to associate confidence in appearance with social ability. A comparison with exploratory item factor analysis highlights our method’s advantages in terms of uncertainty quantification of the clustering process and more detailed and nuanced clustering results. Our method is made available to an existing R package, lsirm12pl (Go et al., 2022). Copyright © 2023, The Authors. All rights reserved.

关键词： Factor analysis

来源：评论

学校读者我要写书评

暂无评论

Constrained Multi-objective Bayesian Optimization through Optimistic Constraints Estimation

arXiv

引用

arXiv 2024年

作者： Li, Diantong Zhang, Fengxue Liu, Chong Chen, Yuxin School of Data Science Chinese University of Hong Kong Shenzhen China Department of Computer Science University of Chicago United States Department of Computer Science University at Albany State University of New York United States

Multi-objective Bayesian optimization has been widely adopted in scientific experiment design, including drug discovery and hyperparameter optimization. In practice, regulatory or safety concerns often impose additional thresholds on certain attributes of the experimental outcomes. Previous work has primarily focused on constrained single-objective optimization tasks or active search under constraints. The existing constrained multi-objective algorithms address the issue with heuristics and approximations, posing challenges to the analysis of the sample efficiency. We propose a novel constrained multi-objective Bayesian optimization algorithm (CMOBO) that balances active learning of the level-set defined on multiple unknowns with multi-objective optimization within the feasible region. We provide both theoretical analysis and empirical evidence, demonstrating the efficacy of our approach on various synthetic benchmarks and real-world applications. © 2024, CC BY.

关键词： Optimization algorithms

来源：评论

学校读者我要写书评

暂无评论

Detection of Speech Signal for Speaker Recognition in Biometric Systems Using Optical Scanning Holography with Double Random Phase Encoding 4

Detection of Speech Signal for Speaker Recognition in Biomet...

引用

4th IEEE International Conference on Mobile Networks and Wireless Communications, ICMNWC 2024

作者： Habelalmateen, Mohammed I. Kalshetty, Jagadevi N. Cindhamani, J. Gayathri, G.M. Jebaseeli, S. Kanmani College of Technical Engineering The Islamic University Department of Computers Techniques Engineering Najaf Iraq Nitte Meenakshi Institute of Technology Department of Computer Science and Engineering Bengaluru India Department of Artificial Intelligence and Data Science Veltech Multitech Dr. Rangarajan Dr. Sakunthala Engineering College Avadi Chennai India Bannari Amman Institute of Technology Department of Humanities Erode India New Prince Shri Bhavani College of Engineering and Technology Department of Information Technology Chennai India

ISBN: (纸本)9798350352931

In recent years, the speaker recognition technology is significant system for securing the biometric authentication and enabling remote logins for telephone banking and enterprise applications. Whereas Speech Signal Detection ensure with accurate speaker identification and verification for restricted access and sensitive data protection. The combination of speaker recognition technology with speech signal detection has accompanied in a new era of biometric authentication, providing unparalleled security and accuracy in verifying identities. However, Vulnerability to voice spoofing attacks leads to imitation of speaker such as mimicry or synthetic speech, compromise system security. To overcome these drawbacks, an Optical Scanning Holography with Double Random Phase Encoding (OSH-DRPE) is proposed using Texas Instruments Massachusetts Institute of Technology (TIMIT) dataset. After collecting the data, the data is preprocessed through normalization approaches for normalize the data. Then, a Mel-Frequency Cepstral Coefficients (MFCC) is employed for feature extraction to extract relevant features, which produces low computational complexity. At last, OSH-DRPE is used as classifier to detect speech signal and speaker recognition with higher accuracy, efficient computation and flexibility. From the results, it clearly shows that the proposed OSH-DRPE obtained higher values in the context of precision, F1 score and accuracy when compared to existing Hidden Markov Model (HMM). © 2024 IEEE.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Certain Investigation of Optimization Methods of Sensor Nodes in Biomedical Recording Systems 9

Certain Investigation of Optimization Methods of Sensor Node...

引用

9th International Conference on Advanced Computing and Communication Systems, ICACCS 2023

作者： Ashokkumar, N. Pappa, C. Kanmani Siva Kumar, A. Srinivasan, Dhamodharan Vijay, M.M. Mohan Babu University Department of ECE Tirupati India Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology Department of ECE Tamilnadu Chennai India SRM Institute of Science and Technology Department of Data Science and Business Systems Chennai India Sri Eshwar College of Engineering Department of Computer and Communication Engineering Tamilnadu Coimbatore India SCAD College of Engineering and Technology Department of ECE Tirunelveli India

ISBN: (纸本)9798350397376

Biosensors will monitor a patient's physiological signals (ECG, EEG, etc.) and send an alarm as soon as irregularities are discovered. Downsized biomedical sensors are the focus of this article, which explores innovative circuit approaches for miniaturised sensors. In order to demonstrate the proposed biomedical compression engine, ECG compression was applied and assessed. A MIP (mixed-integer programming model) is used to represent and solve the clustering issue based on the power consumption analysis. It follows that an architectural model that may be used to design ECG monitoring systems has been provided, as well as an in-depth study of the ECG monitoring system's value chain and a complete evaluation of relevant literature based on an expert classification. Time synchronisation and signal synthesis are the two stages of the distributed cooperative sensing algorithm's execution of ECG signal gathering. This method uses the mutation probability of each solution to prevent local convergence. © 2023 IEEE.

关键词： Electrocardiograms

来源：评论

学校读者我要写书评

暂无评论

Minimum Separator Reconfiguration 18

Minimum Separator Reconfiguration

引用

18th International Symposium on Parameterized and Exact Computation, IPEC 2023

作者： Gomes, Guilherme C. M. Legrand-Duchesne, Clément Mahmoud, Reem Mouawad, Amer E. Okamoto, Yoshio Dos Santos, Vinicius F. Van Der Zanden, Tom C. Department of Computer Science Federal University of Minas Gerais Belo Horizonte Brazil LaBRI CNRS Université de Bordeaux France Virginia Commonwealth University RichmondVA United States Department of Computer Science American University of Beirut Beirut Lebanon Graduate School of Informatics and Engineering The University of Electro-Communications Chofu Japan Department of Data Analytics and Digitalisation Maastricht University Netherlands

ISBN: (纸本)9783959773058

We study the problem of reconfiguring one minimum s-t-separator A into another minimum s-tseparator B in some n-vertex graph G containing two non-adjacent vertices s and t. We consider several variants of the problem as we focus on both the token sliding and token jumping models. Our first contribution is a polynomial-time algorithm that computes (if one exists) a minimum-length sequence of slides transforming A into B. We additionally establish that the existence of a sequence of jumps (which need not be of minimum length) can be decided in polynomial time (by an algorithm that also outputs a witnessing sequence when one exists). In contrast, and somewhat surprisingly, we show that deciding if a sequence of at most jumps can transform A into B is an NP-complete problem. To complement this negative result, we investigate the parameterized complexity of what we believe to be the two most natural parameterized counterparts of the latter problem;in particular, we study the problem of computing a minimum-length sequence of jumps when parameterized by the size k of the minimum s-t-separators and when parameterized by the number of jumps. For the first parameterization, we show that the problem is fixed-parameter tractable, but does not admit a polynomial kernel unless NP ⊆ coNP/poly. We complete the picture by designing a kernel with O(2) vertices and edges for the length of the sequence as a parameter. © 2023 Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing. All rights reserved.

关键词： Separators

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：