检索结果-内蒙古大学图书馆

Statistically characterizing robustness and fidelity of quantum controls and quantum control algorithms

Physical Review A 2023年第3期107卷 032606-032606页

作者： Irtaza Khalid Carrie A. Weidner Edmond A. Jonckheere Sophie G. Shermer Frank C. Langbein School of Computer Science and Informatics Cardiff University Cardiff CF24 4AG United Kingdom Quantum Engineering Technology Laboratories H. H. Wills Physics Laboratory and Department of Electrical and Electronic Engineering University of Bristol Bristol BS8 1FD United Kingdom Department of Electrical and Computer Engineering University of Southern California Los Angeles California 90089 USA Department of Physics Swansea University Swansea SA2 8PP United Kingdom

Robustness of quantum operations or controls is important to build reliable quantum devices. The robustness-infidelity measure (RIMp) is introduced to statistically quantify in a single measure the robustness and fidelity of a controller as the pth order Wasserstein distance between the fidelity distribution of the controller under any uncertainty and an ideal fidelity distribution. The RIMp is the pth root of the pth raw moment of the infidelity distribution. Using a metrization argument, we justify why RIM1 (the average infidelity) is a good practical robustness measure. Based on the RIMp, an algorithmic robustness-infidelity measure (ARIM) is developed to quantify the expected robustness and fidelity of controllers found by a control algorithm. The utility of the RIM and ARIM is demonstrated on energy landscape controllers of spin-12 networks subject to Hamiltonian uncertainty. The robustness and fidelity of individual controllers as well as the expected robustness and fidelity of controllers found by different popular quantum control algorithms are characterized. For algorithm comparisons, stochastic and nonstochastic optimization objectives are considered. Although high fidelity and robustness are often conflicting objectives, some high-fidelity, robust controllers can usually be found, irrespective of the choice of the quantum control algorithm. However, for noisy or stochastic optimization objectives, adaptive sequential decision-making approaches, such as reinforcement learning, have a cost advantage compared to standard control algorithms and, in contrast, the high infidelities obtained are more consistent with high RIM values for low noise levels.

关键词： Machine learning Quantum computation Quantum control Quantum engineering Quantum information theory

来源：评论

学校读者我要写书评

暂无评论

A Semi-Supervised Learning Approach for Fault Detection and Diagnosis in Complex Mechanical Systems

A Semi-Supervised Learning Approach for Fault Detection and ...

引用

IEEE International Conference on Automation science and engineering (CASE)

作者： Bahman Askari Graziana Cavone Raffaele Carli Antoine Grall Mariagrazia Dotoli Department of Electrical and Information Engineering of the Polytechnic of Bari Italy Department of Civil Computer Science and Aeronautical Technologies Engineering of the University Roma Tre Italy LIST3N-LM2S University of Technology of Troyes France

The integration of artificial intelligence in mechanical fault detection and diagnosis (FDD) helps to increase reliability, reduce costs, and improve the overall performance of mechanical systems in Industry 4.0 applications. Most interesting industrial applications nowadays come from dynamic environments where data are generated continuously over time and where the labeled data are scarce and expensive. Therefore, semi-supervised learning (SSL) can be particularly useful in FDD because faults may be rare or difficult to identify, and may not be fully represented in the labeled data. By using a combination of labeled and unlabeled data, SSL can help to identify these rare or difficult-to-detect faults, leading to more effective FDD. In this paper, graph-based SSL relying on label propagation is combined with conventional classification algorithms to detect potential failures in complex mechanical systems. Experimental results on realistic pneumatic and hydraulic systems from the related literature show that the proposed method can effectively enlarge the labeled datasets and interestingly identify different types of non-nominal conditions with higher accuracy compared to baseline methodologies.

关键词：

来源：评论

学校读者我要写书评

暂无评论

MindGPT: Advancing Human-AI Interaction with Non-Invasive fNIRS-Based Imagined Speech Decoding

arXiv

引用

arXiv 2024年

作者： Zhang, Suyi Alam, Ekram Baber, Jack Bianca, Francesca Turner, Edward Chamanzar, Maysam Dehghani, Hamid MindPortal Ltd LondonEC2A 3PG United Kingdom Department of Electrical and Computer Engineering Carnegie Mellon University PittsburghPA15213 United States Department of Biomedical Engineering Carnegie Mellon University PittsburghPA15213 United States Neuroscience Institute Carnegie Mellon University PittsburghPA15213 United States School of Computer Science University of Birmingham BirminghamB15 2TT United Kingdom

In the coming decade, artificial intelligence systems are set to revolutionise every industry and facet of human life. Building communication systems that enable seamless and symbiotic communication between humans and AI agents is increasingly important. This research advances the field of human-AI interaction by developing an innovative approach to decode imagined speech using non-invasive high-density functional near-infrared spectroscopy (fNIRS). Notably, this study introduces MindGPT, the first thought-to-LLM (large language model) system in the world. This study focuses on enhancing human-AI communication by utilising fNIRS data to develop a proprietary AI model called MindGPT capable of decoding imagined speech. Hemodynamic responses representing neural activity were collected from four participants instructed to imagine three different sentences. An Extra Trees Classifier (XTC) model was employed to decode neural patterns and differentiate imagined speech from rest conditions, achieving an average accuracy of ~66% across participants, with the best average accuracy at 71%. To further decode neural signals associated with specific imagined sentences, a convolutional neural network (CNN) and a ridge regression model were used as decoders. The CNN model demonstrated an advantage with minimally preprocessed optical density data, outperforming the ridge regression model in this task. Our results showed significant decoding accuracy for imagined speech, with the ridge regression model achieving a best accuracy of 57% for one participant (chance level: 33%, p-value © 2024, CC BY-NC-ND.

关键词： Near infrared spectroscopy

来源：评论

学校读者我要写书评

暂无评论

An Accurate Path Tracking Algorithm for Autonomous Vehicles Based on Pure Pursuit with Systematically Designed Look-Ahead Distance and Sideslip Compensation

An Accurate Path Tracking Algorithm for Autonomous Vehicles ...

引用

International Conference on Control, Automation and Systems ( ICCAS)

作者： Ji Hwan Seo Sung Hoon Youn Jungeun Kim Kyoung-Dae Kim Department of Electrical Engineering and Computer Science DGIST Daegu Korea Convergence Technology Development Team Autonomous a2z Anyang Korea Development Team Vueron Technology Seoul Korea

Pure pursuit is still being utilized as a path tracking algorithm for autonomous vehicles due to its advantages, such as robustness and easy implementation. However, using an inappropriate look-ahead distance and ignoring the effects of dynamics deteriorate the path tracking performance of pure pursuit. To improve the tracking performance, we propose a pure pursuit-based path tracking algorithm that quickly reduces tracking error and considers the sideslip by dynamics. For fast tracking error reduction, a dynamic look-ahead distance is designed as a function of speed and path curvature by analyzing the tracking error change according to the look-ahead distance. In addition, to compensate for the sideslip effect, the desired steering angle is calculated using the sideslip integrated-pure pursuit geometry. Simulation results show that the proposed algorithm significantly reduced the tracking error compared to the classical pure pursuit at a maximum 70 km/h speed. Also, the proposed method outperformed existing improved pure pursuits.

关键词：

来源：评论

学校读者我要写书评

暂无评论

TIMEPREST: TIME AND MEMORY EFFICIENT PIPELINE PARALLEL DNN TRAINING WITH REMOVED STALENESS

arXiv

引用

arXiv 2024年

作者： Dutta, Ankita Chaki, Nabendu De, Rajat K. Machine Intelligence Unit Indian Statistical Institute 203 Barrackpore Trunk Road Kolkata700108 India Department of Computer Science and Engineering University of Calcutta JD-2 Sector III Salt Lake City Kolkata700098 India

DNN training is extremely time-consuming, necessitating efficient multi-accelerator parallelization, where a single iteration of training is split over the available accelerators. Current approaches often parallelize training by using intra-batch parallelization. Combining inter-batch pipeline parallelism with intra-batch parallelism is a common approach to further improve parallel training throughput. In this article, we develop a system, called TiMePReSt, that adds both of them, but in a different way which helps to better overlap computation and communication within a mini-batch, and limits the amount of communication. The traditional pipeline parallel training of DNNs maintains similar working principle as sequential or conventional training of DNNs. Thus, it suffers from high GPU memory footprint during training to maintain consistent version of weights in forward and backward passes of a mini-batch, similar to sequential ***, it has been shown experimentally that violating consistency of weight versions does not necessarily reduce prediction capability of a parallely trained DNN. TiMePReSt helps to overcome GPU memory overhead and achieve zero degree of staleness of weights by sacrificing sequential stability, but not effecting prediction capability. State-of-the-art pipeline parallel DNN training techniques often become costly in terms of training time. In order to address this issue, TiMePReSt introduces a variant of intra-batch parallelism that parallelizes the forward pass of each mini-batch by decomposing it into smaller micro-batches. Moreover, synchronization between backward and forward passes are performed in a novel way reduce training time in TiMePReSt. The chances of occurring multiple sequence problem and its relation with version difference have been observed in TiMePReSt. The version difference is observed to be varied with the number of micro-batches and worker machines. In this paper, a mathematical relationship between the number of

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Hybrid Quantum Noise Approximation and Pattern Analysis on Parameterized Component Distributions

arXiv

引用

arXiv 2024年

作者： Chakraborty, Mouli Mukherjee, Anshu Krikidis, Ioannis Nag, Avishek Chandra, Subhash School of Natural Sciences Trinity College Dublin The University of Dublin College Green Dublin 2 Ireland School of Electrical and Electronic Engineering University College Dublin Belfield Dublin 4 Ireland Department of Electrical and Computer Engineering IRIDA Research Centre for Communication Technologies University of Cyprus Nicosia Cyprus School of Computer Science University College Dublin Belfield Dublin 4 Ireland

Noise is a vital factor in determining the accuracy of processing the information of the quantum channel. One must consider classical noise effects associated with quantum noise sources for more realistic modelling of quantum channels. A hybrid quantum noise model incorporating both quantum Poisson noise and classical additive white Gaussian noise (AWGN) can be interpreted as an infinite mixture of Gaussians with weightage from the Poisson distribution. The entropy measure of this function is difficult to calculate. This research developed how the infinite mixture can be well approximated by a finite mixture distribution depending on the Poisson parametric setting compared to the number of mixture components. The mathematical analysis of the characterization of hybrid quantum noise has been demonstrated based on Gaussian and Poisson parametric analysis. This helps in the pattern analysis of the parametric values of the component distribution, and it also helps in the calculation of hybrid noise entropy to understand hybrid quantum noise better. © 2024, CC BY.

关键词： Quantum noise

来源：评论

学校读者我要写书评

暂无评论

Response to an applied electric field in an antiferroelectric 1/2 subphase: The role of thermal fluctuations

引用

Physical Review E 2023年第6期107卷 064701-064701页

作者： Koichiro Shirota Atsuo Fukuda Neelam Yadav Vitaly P. Panov Jagdish K. Vij Yutaka Yamagata Ken Ishikawa RIKEN Center for Advanced Photonics (RAP) 2-1 Hirosawa Wako Saitama 351-0198 Japan Department of Electronic and Electrical Engineering Trinity College Dublin The University of Dublin Dublin 2 Ireland Department of Electronic and Computer Engineering Sungkyunkwan University Suwon Gyeonggi-do 16419 Republic of Korea Department of Materials Science and Engineering School of Materials and Chemical Technology Tokyo Institute of Technology 2-12-1 S8-28 Ookayama Meguro-ku Tokyo 152-8552 Japan

The response to an applied electric field in the qT=1/2 subphase of the MC881-MC452 binary mixture system is studied by using thick homeotropically aligned cells. In the ordinary antiferroelectric SmCA* and 1/2 (sub)phases, some nonplanar asymmetric distortions in the antiferroelectric unit cell structure produce induced polarization in the applied field direction, starts to unwind the helix from the beginning, and tends to align the averaged tilt plane direction parallel to the applied field. In the 1/2 subphase under consideration, however, the helix resists being deformed at the beginning and then the thresholdlike steep increase of birefringence Δn occurs in the transition from 1/2 to unwound SmC* at a field of less than 0.5 V/µm; we conclude that the thermal fluctuations play an important role in promoting the director flip-flopping in a single layer under the applied field and bring about additional induced polarization, which counteracts the aforementioned ordinary induced one and prevents the helix from unwinding. This suggests that the Langevin-like director reorientation is the mechanism of the V-shaped switching which was actually observed in the thin films of Mitsui mixture [Phys. Rev. Lett. 87, 015701 (2001)] and must have been used in prototyped thresholdless antiferroelectric liquid-crystal displays.

关键词： Ferroelectric liquid crystals Liquid crystals Smectic liquid crystals

来源：评论

学校读者我要写书评

暂无评论

Privacy-aware authentication scheme for electric vehicle in-motion wireless charging

Privacy-aware authentication scheme for electric vehicle in-...

引用

2020 International Symposium on Networks, computers and Communications, ISNCC 2020

作者： Hamouid, Khaled Adi, Kamel University of Batna 2 LaSTIC Research Laboratory Department of Computer Science Batna Algeria University of Quebec in Outaouais Computer Security Research Laboratory Department of Computer Science and Engineering Quebec Canada

ISBN: (纸本)9781728156286

In-motion wireless charging services have emerged for the development of Electric Vehicles (EVs) industry. Fast authentication and privacy awareness are the major concerns in this promising technology. This paper proposes FLPA, a fast and anonymous authentication scheme for EVs charging on the move, which preserves the identity and location privacy of EVs during the recharging process. In order to deal with EV's high mobility and constrained resources of charging pads, FLPA provides lightweight and fast EVs authentication to charging pads while ensuring secure and fair-billing based on authenticated pairwise keys and coin chains. Through a performance analysis, we demonstrate our scheme's advantage compared to current solutions. © 2020 IEEE.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Hybrid Quantum Noise Approximation and Pattern Analysis on Parameterized Component Distributions

Hybrid Quantum Noise Approximation and Pattern Analysis on P...

引用

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

作者： Mouli Chakraborty Anshu Mukherjee Ioannis Krikidis Avishek Nag Subhash Chandra School of Natural Sciences Trinity College Dublin The University of Dublin College Green Dublin 2 Ireland School of Electrical and Electronic Engineering University College Dublin Belfield Dublin 4 Ireland Department of Electrical and Computer Engineering IRIDA Research Centre for Communication Technologies University of Cyprus Nicosia Cyprus School of Computer Science University College Dublin Belfield Dublin 4 Ireland

ISBN: (数字)9798350376715

ISBN: (纸本)9798350376722

关键词： Analytical models Correlation Additives Accuracy Noise Quantum channels Entropy Noise measurement Pattern analysis Photonics

来源：评论

学校读者我要写书评

暂无评论

Generation of All Rooted Ordered Trees

SSRN

引用

SSRN 2022年

作者： Chakraborty, Sumit Bhattacharyya, Riddhasree Chakraborty, Maumita Pal, Rajat Kumar Department of Computer Science Ramsaday College Amta Howrah711401 India Department of Computer Science and Engineering University of Calcutta JD-2 Sector III Salt Lake Kolkata700106 India Department of Computer Science and Technology University of Engineering and Management Kolkata University Area Plot No. III – B/5 New Town Action Area – III Kolkata700160 India

Enumeration of Rooted trees is a major area of research in computer science and mathematics. Rooted ordered trees are used to solve combinatorial optimization problems, analyze the structure of chemical compounds, and many others. Several researchers from different continents have given different algorithms for generating rooted ordered trees. Different encodings have been used to represent the rooted ordered trees by these researchers. We have introduced the children sequence to represent each rooted ordered tree. In this paper, our proposed algorithm, All-Ordered-Trees , uses an efficient algebraic approach to generate all the children sequences (representing rooted ordered trees) in lexicographic order for a given number of vertices, n . With slight modification, this algorithm can also be used to generate degree bound rooted ordered trees. The space complexity of the algorithm, All-Ordered-Trees , is O ( n ) and the number of operations required to generate a tree is approximately bounded by n /3. We have also implemented three existing algorithms in this domain and included the experimental results of these algorithms with the experimental results of our proposed algorithm for comparative study. From the experimental results, we observe that each tree can be generated in approximately constant time. © 2022, The Authors. All rights reserved.

关键词： Chemical analysis

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：