检索结果-内蒙古大学图书馆

algorithms That Can Deny Care, and a Call for AI Explainability

COMPUTER 2024年第7期57卷 109-112页

作者： Talia, Domenico Univ Calabria Comp Engn I-87036 Arcavacata Di Rende Italy

Recent examples of negative use of big data and machine learning come from their use in health care decisions involving a large number of patients. These cases highlight the need for algorithm explainability to help us better understand how artificial intelligence works in solving problems so we can then evaluate its accuracy and effectiveness.

关键词： Intelligence Machine Learning Learning algorithms Big Data Use Of Machine Learning Use Of Big Data Class Action Lawsuit Health Insurance Black Box prediction algorithms General Data Protection Regulation Intelligence Technology Social Profiles Algorithm Works Post Acute Care Algorithm In Scenarios

来源：评论

学校读者我要写书评

暂无评论

A COMPARISON OF 2 ACTIVELY COMPENSATED ACTIVE-RC FILTERS

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS 1990年第2期37卷 296-299页

作者： NATARAJAN, S Department of Electrical Engineering Tennessee Technological University Cookeville TN USA

A number of authors have proposed the use of additional operational amplifiers (op amps) for compensating the detrimental effects of the frequency-dependent gains of the op amps already used in the circuit. This technique is known as active compensation. Among all such actively compensated circuits, two such circuits have been derived from the single amplifier biquad (SAB) of Deliyannis-Friend. These two circuits possess very low passive and active sensitivities. The purpose of this work is to compare their performances not only with respect to the active sensitivities but also from the point of requiring matched op amps and their stability.< >

关键词： Operational amplifiers Filters Polynomials Circuit stability Circuits and systems Control theory Linear systems Circuit theory Stochastic processes prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

On the Skewness of the LMS Adaptive Weights

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS 2021年第8期68卷 3022-3026页

作者： Silva, Thiago T. P. Igreja, Filipe Lara, Pedro Tarrataca, Luis Kar, Asutosh Haddad, Diego B. Ctr Fed Educ Tecnol Celso Suckow da Fonseca Dept Comp Engn BR-20271204 Rio De Janeiro Brazil Indian Inst Informat Technol Design & Mfg Elect & Comp Engn Dept Chennai 600127 Tamil Nadu India

The adjustable weights of adaptive filtering algorithms are usually assumed to obey a Gaussian distribution. This is somewhat natural under maximal-entropy considerations, since most analyses in the open literature only take into account first-and second-order statistics. This work investigates the third-order statistical feature known as skewness of the least mean square parameters distribution. Two theoretical analyses for skewness estimation are proposed: i) one that employs the independence assumption, which states that the excitation data is statistically independent from the adaptive weights;ii) one derived from the exact expectation analysis, a method that is able to predict the learning capabilities of the least mean square algorithm even when the step size is not infinitesimally small. This brief shows that the skewness of the adaptive weights distribution may present a large deviation from the common Gaussian assumption, especially in the first phase of the learning. Furthermore, it is also demonstrated that the skewness may grow without limit even when adaptive weights present convergence in both average and mean square behaviors.

关键词： Mathematical model Adaptation models Stochastic processes Stability analysis prediction algorithms Estimation Analytical models Adaptive filtering exact expectation analysis LMS convergence analysis skewness

来源：评论

学校读者我要写书评

暂无评论

Risk Assessment and Mitigation in Local Path Planning for Autonomous Vehicles With LSTM Based Predictive Model

引用

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING 2022年第4期19卷 2738-2749页

作者： Wang, Hong Lu, Bing Li, Jun Liu, Teng Xing, Yang Lv, Chen Cao, Dongpu Li, Jingxuan Zhang, Jinwei Hashemi, Ehsan Tsinghua Univ Tsinghua Intelligent Vehicle Design & Safety Res Beijing 100084 Peoples R China Beijing Inst Technol Beijing 100811 Peoples R China Waterloo Univ Dept Mech & Mechatron Engn Waterloo ON N2L 3G1 Canada Nanyang Technol Univ Sch Mech & Aerosp Engn Singapore Singapore Nanjing Univ Aeronaut & Astronaut Coll Energy & Power Engn Nanjing 210016 Peoples R China

Accurate trajectory prediction of surrounding vehicles enables lower risk path planning in advance for autonomous vehicles, thus promising the safety of automated driving. A low-risk and high-efficiency path planning approach is proposed for autonomous driving based on the high-performance and practical trajectory prediction method. A long short-term memory (LSTM) network is trained and tested using the highD dataset, and the validated LSTM is used to predict the trajectories of surrounding vehicles combining the information extracted from vehicle-to-vehicle (V2V) technology. A risk assessment and mitigation-based local path planning algorithm is proposed according to the information of predicted trajectories of surrounding vehicles. Two driving scenarios are extracted and reconstructed from the highD dataset for validation and evaluation, i.e., an active lane-change scenario and a longitudinal collision-avoidance scenario. The results illustrate that the risk is mitigated and the driving efficiency is improved with the proposed path planning algorithm comparing to the constant-velocity prediction and the prediction method of the nonlinear input-output (NIO) network, especially when the velocity and trajectory with sudden changes.

关键词： Safety Risk management Trajectory Tires prediction algorithms Accidents Predictive models HighD dataset local path planning long short-term memory (LSTM) model predictive control (MPC) risk assessment risk mitigation

来源：评论

学校读者我要写书评

暂无评论

Customer-Centered Pricing Strategy Based on Privacy-Preserving Load Disaggregation

引用

IEEE TRANSACTIONS ON SMART GRID 2023年第5期14卷 3401-3412页

作者： Tao, Yuechuan Qiu, Jing Lai, Shuying Sun, Xianzhuo Ma, Yuan Zhao, Junhua Univ Sydney Sch Elect & Informat Engn Sydney NSW 2006 Australia Chinese Univ Hong Kong Shenzhen Sch Sci & Engn Shenzhen 518100 Peoples R China Chinese Univ Hong Kong Shenzhen Shenzhen Inst Artificial Intelligence & Robot Soc Shenzhen 518100 Peoples R China

Demand response (DR) is a demand reduction or shift of electricity use by customers to make electricity systems flexible and reliable, which is beneficial under increasing shares of intermittent renewable energy. For residential loads, thermostatically controlled loads (TCLs) are considered as major DR resources. In a price-based DR program, an aggregation agent, such as a retailer, formulates price signals to stimulate the customers to change electricity usage patterns. The conventional DR management methods fully rely on mathematical models to describe the customer ’s price responsiveness. However, it is difficult to fully master the customers ’ detailed demand elasticities, cost functions, and utility functions in practice. Hence, in this paper, we proposed a data-driven non-intrusive load monitoring (NILM) approach to study the customers ’ power consumption behaviors and usage characteristics. Based on NILM, the DR potential of the TCLs can be properly estimated, which assists the retailer in formulating a proper pricing strategy. To realize privacy protection, a privacy-preserving NILM algorithm is proposed. The proposed methodologies are verified in case studies. It is concluded that the proposed NILM algorithm not only reaches a better prediction performance than state-of-art works but also can protect privacy by slightly sacrificing accuracy. The DR pricing strategy with NILM integrated brings more profit and lower risks to the retailer, whose results are close to the fully model-based method with strong assumptions. Furthermore, a NILM algorithm with higher performance can help the retailer earn more benefits and help the grids better realize DR requirements.

关键词： Signal processing algorithms Pricing prediction algorithms Power demand Elasticity Behavioral sciences Smart meters Demand response potential pricing strategy thermostatically controlled loads non-intrusive load monitoring privacy protection

来源：评论

学校读者我要写书评

暂无评论

Advancing High-Mountain Precipitation Reconstruction Through Merging of Multiple Data Sources: Triple Collocation Versus Signal-to-Noise Ratio Optimization

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2024年 62卷

作者： Shah, Suraj Liu, Yi Kim, Seokhyeon Sharma, Ashish Univ New South Wales Sch Civil & Environm Engn Sydney NSW 2052 Australia Kyung Hee Univ Coll Engn Dept Civil Engn Yongin 17104 South Korea

Merging multisource satellite precipitation data through triple collocation (TC) is an established method for enhancing precipitation estimates without ground observations. However, the effectiveness of TC is limited by three constraints: the need for three data sources, the sum of weights equalling one, and no error cross correlation (ECC) among datasets. These limitations restrict applications to areas with three distinct data sources, may result in suboptimal combinations, and can introduce biases. We explored a new framework, signal-to-noise ratio optimization (SNR-opt), which alleviates these restrictions. Applying TC and SNR-opt to datasets in the Himalayas, including three satellite-based (IMERG Final, CMORPH, and PERSIANN), one soil moisture-based (SM2RAIN-ASCAT), and one reanalysis dataset (ERA5), we assessed their performance. The flexibility of SNR-opt allowed for creating products suited to varying rainfall conditions and reconstructing extreme precipitation products for high-mountain Asia (HMA). The results show that: 1) the data triplet containing IMERG-final was the best performing among all merging products;2) the merging method significantly enhanced dataset accuracy compared to the parent datasets, with SNR-opt proving superior to TC in all selected precipitation indices evaluated;and 3) SNR-opt had the flexibility of estimating both low and heavy rainfall, with a reduced bias for extreme precipitation, using fewer inputs (two) compared to TC. These advantages were found to be especially notable for the high-mountain region in focus, given the significant variability and limited observational data available.

关键词： Precipitation Merging Satellites Meteorology Soil moisture Rain Asia Accuracy Soft sensors prediction algorithms Extreme precipitation high-mountain Asia (HMA) satellite precipitation merging signal-to-noise ratio optimization (SNR-opt) triple collocation (TC)

来源：评论

学校读者我要写书评

暂无评论

Imbalanced Sample Selection With Deep Reinforcement Learning for Fault Diagnosis

引用

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 2022年第4期18卷 2518-2527页

作者： Fan, Saite Zhang, Xinmin Song, Zhihuan Zhejiang Univ State Key Lab Ind Control Technol Coll Control Sci & Engn Hangzhou 310027 Zhejiang Peoples R China Collaborat Innovat Ctr Artificial Intelligence MO Hangzhou 310027 Peoples R China

An imbalanced number of faulty and normal samples causes serious damage to the performance of the conventional diagnosis methods. To settle the data-imbalance fault diagnosis problem, this article presents a novel general imbalanced sample selection strategy (DiagSelect) based on deep reinforcement learning. In DiagSelect, the problem of imbalanced sample selection from the training set is formulated as a multiarmed bandit problem of deep reinforcement learning. The nondifferentiable optimization problem of imbalanced sample selection can be solved by the Markov decision process. The parameters of DiagSelect can be optimized by REINFORCE with the feedback of the validation set. DiagSelect performs intelligent imbalanced sample selection to obtain better diagnosis performance autonomously. As a data-level technique, DiagSelect can be easily used in conjunction with the conventional diagnosis models. DiagSelect is validated in a synthetic dataset and an industrial process dataset. The results have shown the effectiveness, stability, and transferability of DiagSelect.

关键词： Fault diagnosis Training Adaptation models Reinforcement learning Testing Task analysis prediction algorithms Deep reinforcement learning fault diagnosis imbalanced sample selection industrial process monitoring multi-armed bandit (MAB) problem nondifferentiable optimization problem

来源：评论

学校读者我要写书评

暂无评论

Special Issue on Big Data Networking-Challenges and Applications

引用

JOURNAL OF COMMUNICATIONS AND NETWORKS 2015年第6期17卷 545-548页

作者： Han, Zhu Bennis, Mehdi Wang, Dan Kwon, Taekyoung Cui, Shuguang Univ Maryland College Pk MD 20742 USA Boise State Univ Boise ID 83725 USA Univ Oulu CWC SF-90100 Oulu Finland Seoul Natl Univ Dept Comp Sci & Engn Seoul 151 South Korea Univ Calif Los Angeles Los Angeles CA 90024 USA CUNY New York NY 10016 USA Texas A&M Univ Elect & Comp Engn College Stn TX USA

Big data is more than a matter of size; it is an emerging paradigm of data of very large size (volume) and fast in/out (velocity), from various sources (variety), and of high value for knowledge extraction and decision making. Technological advances in data gathering have led to a rapid proliferation of big data in diverse areas such as remote sensing, medicine, the Internet, and social sectors. Such data brings opportunities and challenges to scientists and engineers. In order for us to make use of this massive amount of data, new data management and computational approaches are needed to permit scientists and engineers to analyze the data in (nearly) real time, often in a distributed or streaming manner. Various technologies are being discussed, and some have been realized, to support the handling of big data. In addition, big dataset cannot be stored in one location, and massively parallel processing databases and scalable storage systems are being designed to store the large datasets. What is more, big data generates an industry of supporting architectures; many cloud computing platforms and frameworks are developed to handle the big data operations, such as MapReduce. To deal with different properties of big data, different algorithms also need to be developed. Overall, big data is an opportunity to find insights in new and emerging types of data and content, to make models more agile, and to answer questions that were previously considered beyond our reach. The purpose of this special is to highlight some recent advancement to address such challenges in the big data era.

关键词： Big data Wireless communication Distributed databases Data mining Algorithm design and analysis prediction algorithms Mobile communication

来源：评论

学校读者我要写书评

暂无评论

Machine Learning in the Assessment of Meat Quality

引用

IT PROFESSIONAL 2020年第3期22卷 39-41页

作者： Penning, Bryan W. Snelling, Warren M. Woodward-Greene, M. Jennifer ARS USDA Wooster OH 44691 USA ARS USDA Clay Ctr NE USA ARS USDA Natl Ag Lib Indexing & Informat Branch Baltimore MD USA

We compare two approaches to automate carcass quality grading using different artificial intelligence methods. The first is based on image analysis, and the second uses state-of-the-art Rapid Evaporative Ionization Mass Spectrometry. Both employ machine learning (ML) to increase the speed and accuracy of carcass quality evaluation. The image analysis method increased speed and accuracy for all quality measures except marbling when compared to human meat inspectors. The mass spectrometry method tested eight ML algorithms, and achieved an impressive 81.5% to 99% accuracy in predicting carcass quality traits. However, this accuracy was dependent on the trait examined, so ML algorithms were not the answer for all traits.

关键词： prediction algorithms Mass spectroscopy Image analysis Muscles Cameras Production Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

A Sharp Analysis of Covariate Adjusted Precision Matrix Estimation via Alternating Projected Gradient Descent

引用

IEEE SIGNAL PROCESSING LETTERS 2022年 29卷 877-881页

作者： Lv, Xiao Cui, Wei Liu, Yulong Beijing Inst Technol Sch Informat & Elect Beijing 100081 Peoples R China Beijing Inst Technol Sch Phys Beijing 100081 Peoples R China

In this letter, we present a sharp algorithmic analysis for alternating projected gradient descent which is used to solve the covariate adjusted precision matrix estimation problem in high-dimensional settings. By introducing a new analytical tool (the generic chaining), we remove the impractical resampling assumption used in the literature. The new analysis also demonstrates that this algorithm not only enjoys a linear convergence rate in the absence of convexity, but also attains the minimax rate with optimal order of sample complexity. Our results, meanwhile, reveal a time-data tradeoff in this problem. Numerical experiments are provided to verify our theoretical results.

关键词： Signal processing algorithms Convergence Estimation Complexity theory Estimation error Linear regression prediction algorithms Multivariate linear regression covariate adjusted precision matrix estimation alternating gradient descent

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：