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 483-565页

作者： Jayakumar Subramanian Amit Sinha Raihan Seraj Aditya Mahajan Media and Data Science Research Lab Digital Experience Cloud Adobe Systems India Private Limited Noida Uttar Pradesh India Department of Electrical and Computer Engineering McGill University Montreal Quebec Canada

We propose a theoretical framework for approximate planning and learning in partially observed systems. Our framework is based on the fundamental notion of information state. We provide two definitions of information state--i) a function of history which is sufficient to compute the expected reward and predict its next value; ii) a function of the history which can be recursively updated and is sufficient to compute the expected reward and predict the next observation. An information state always leads to a dynamic programming decomposition. Our key result is to show that if a function of the history (called approximate information state (AIS)) approximately satisfies the properties of the information state, then there is a corresponding approximate dynamic program. We show that the policy computed using this is approximately optimal with bounded loss of optimality. We show that several approximations in state, observation and action spaces in literature can be viewed as instances of AIS. In some of these cases, we obtain tighter bounds. A salient feature of AIS is that it can be learnt from data. We present AIS based multi-time scale policy gradient algorithms and detailed numerical experiments with low, moderate and high dimensional environments.

关键词： partially observed reinforcement learning partially observable Markov decision processes approximate dynamic programming information state approximate information state

来源：评论

学校读者我要写书评

暂无评论

Bias sputtering of granular L10-FePt films with hexagonal boron nitride grain boundaries

arXiv

引用

arXiv 2023年

作者： Xu, Chengchao Varaprasad, B.S.D.Ch.S. Laughlin, David E. Zhu, Jian-Gang Data Storage Systems Center Carnegie Mellon University PittsburghPA15213 United States Electrical and Computer Engineering Department Carnegie Mellon University PittsburghPA15213 United States Materials Science and Engineering Department Carnegie Mellon University PittsburghPA15213 United States

In this paper, we present an experimental study of L10-FePt granular films with crystalline/amorphous boron nitride (BN) grain boundary materials for heat assisted magnetic recording (HAMR). It is found that an adequate RF substrate bias yields the formation of hexagonal boron nitride (h-BN) nanosheets in grain boundaries, facilitating the columnar growth of FePt grains during sputtering at high temperatures. The h-BN monolayers conform to the side surfaces of columnar FePt grains, completely encircling individual FePt grains. The resulting core-shell FePt/h-BN nanostructures appear to be highly promising for HAMR application. The high thermal stability of h-BN grain boundaries allows the deposition temperature to be as high as 800oC such that high order parameters of FePt L10 phase have been obtained. For the fabricated FePt/h-BN thin film, excellent granular microstructure with FePt grains of 6.5nm in diameter and 11.5nm in height has been achieved along with good magnetic hysteresis properties. © 2023, CC BY.

关键词： Grain boundaries

来源：评论

学校读者我要写书评

暂无评论

A Review of Algorithms’s Complexities on Different Valued Sorted and Unsorted Data

A Review of Algorithms’s Complexities on Different Valued S...

引用

IT and Industrial Technologies (ICIT), International Conference on

作者： Attia Shabbir Asad Majeed Mahnoor Iftikhar Raja Hashim Ali Usama Arshad Muhammad Zeeshan Shabbir Ali Zeeshan Ijaz Nisar Ali Ali Aftab Faculty of Computer Science & Engineering GIK Institute of Engg. Sciences & Technology Topi Pakistan Dept. of Technology & Software Engineering University of Europe for Applied Sciences Berlin Germany Department of Electrical Engineering University of Engineering & Technology (UET) Lahore Pakistan Faculty of Electronic Systems Engineering University of Regina Regina Canada

Sorting and searching are critical processes for effecttive data analysis. In this paper, we evaluate the performance of various sorting and searching algorithms and compare their time and space complexities on both sorted and unsorted data. The algorithms we analyzed include six common sorting algorithms (insertion, radix, bucket, merge, bubble, and quick sort) and three search algorithms (linear, binary, and jump search). The results of our study provide insights into the best algorithms to use for different input sizes and types of data. It was found that for small input sizes, all algorithms perform similarly, but for larger input sizes, insertion and radix sorts are better for time complexity while bubble sort is better for space complexity. Additionally, jump search outperformed linear and binary search algorithms in both time and space complexity. Besides, difference between time and space complexity of sorted and unsorted data was significant.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Measurement of paracetamol concentration using a fiber laser system

Measurement of paracetamol concentration using a fiber laser...

引用

Bioengineering (ENBENG), Portuguese Meeting in

作者： Liliana Soares Rosa Ana Perez-Herrera Susana Novais António Ferreira Orlando Frazão Susana Silva Faculty of Engineering INESC TEC - Institute for Systems and Computer Engineering Technology and Science University of Porto Porto Portugal Department of Electrical Electronic and Communication Engineering Public Univ. of Navarra ISC - Institute of Smart Cities Public Univ. of Navarra Pamplona Spain INESC TEC - Institute for Systems and Computer Engineering Technology and Science Porto Portugal LEPABE-Laboratory for Process Engineering Environment Biotechnology and Energy University of Porto Porto Portugal ALiCE - Associate Laboratory in Chemical Engineering University of Porto Porto Portugal

A linear fiber laser system for measurements of paracetamol concentration is experimentally demonstrated. The cavity is based on a fiber loop mirror and an FBG centered at 1567.8 nm. The sensing head corresponds to a refractometric sensor, whose which principle of operation is based on Fresnel reflection in the fiber tip (FBG side). The system works at detected variations of paracetamol concentrations with a sensitivity of $[(8.74\pm 0.34)\times 10^{-5}]\ \ \mu \mathrm{W}/(\mathrm{g}/\text{kg})$ and a resolution of 2.77 g/kg. The results prove that the fiber laser system could be an asset for processing industries, specifically for non-invasive and real-time measurements of concentration.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Design in STAR-RIS Assisted Communication System with Antenna Selection

Energy-Efficient Design in STAR-RIS Assisted Communication S...

引用

IEEE Conference on Global Communications (GLOBECOM)

作者： Guangyang Zhang Yang Lu Bo Ai Zhangdui Zhong Zhiguo Ding Tony Q. S. Quek School of Electronic and Information Engineering Beijing Jiaotong University Beijing China School of Computer and Information Technology Beijing Jiaotong University Beijing China Key Laboratory of Railway Industry of Broadband Mobile Information Communications Beijing Jiaotong University Beijing China Department of Electrical Engineering and Computer Science Khalifa University Abu Dhabi UAE Department of Electrical and Electronic Engineering University of Manchester Manchester UK Information Systems Technology and Design Pillar Singapore University of Technology and Design Singapore

This paper investigates the energy-efficient beamforming design in a simultaneous transmission and reflection-reconfigurable intelligent surface (STAR-RIS) assisted wireless communication system, where the antenna selection scheme is adopted. An energy efficiency (EE) maximization problem is formulated by optimizing the transmit beamformers and the phase shift vectors subject to the power budget constraint of the base station (BS), the maximum transmit power constraint per antenna and the users' data rate requirements. An alternating optimization-based algorithm is proposed to tackle the coupled variables, and the quadratic transform is used to deal with the fractional formulations. Simulation results demonstrate that the antenna selection scheme can significantly improve the EE performance by suppressing the energy consumption due to massive antennas. With the assistance of the STAR-RIS, the EE performance is further enhanced.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Intelligent Recognition of Immune Cells: Analysis and Research

Intelligent Recognition of Immune Cells: Analysis and Resear...

引用

IEEE International Conference on e-Business engineering (ICEBE)

作者： Chih-En Li Wei-Lun Liu Kun-Yi Lai Yu-Hsiang Chiang Xiaowei Yan Hao-Xiang Chen Hsu-Ching Li Wan-Shu Cheng Department of Electrical Engineering National Kaohsiung University of Science and Technology Kaohsiung Taiwan Institute for Systems Biology Seattle Washington U.S.A Dep. of Computer Science and Information Management Providence University Taichung Taiwan

Immunotherapy utilizes the potential of the patient's immune system to fight disease. T-lymphocytes have become the central focus of the immune system in the fight against cancer. One of the fundamental studies in immunotherapy is how to accurately identify various types of immune cells. Therefore, the study of immune-related cells will be an important reference for the treatment of COVID19 or other related diseases. In this paper, a series of experiments were conducted with COVID-19 patients as the starting point to design a strategy to identify the important subtypes of CD$4^{+}$T cells. In this paper, an intelligent machine learning algorithm is proposed according to the results of different experiments of linear and nonlinear dimension reduction algorithms, integrating the important characteristics of these two algorithms and improving the performance. Our experimental results show that the algorithm designed based on the front-end experiments can effectively distinguish the important lymphocytes. The proposed algorithm not only improves the performance, but also the dimension reduction procedure keeps the important features of both global and regional data. The experimental results of this paper will be beneficial for the subsequent analysis of the clustering in identifying important subtypes of CD$4^{+}$T cells. Most importantly, this wm significantly improve the procedure of manually integrating and labeling the recorded data after several tedious and monotonous analyses with external software.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Convergence Analysis of the Best Response Algorithm for Time-Varying Games

Convergence Analysis of the Best Response Algorithm for Time...

引用

IEEE Conference on Decision and Control

作者： Zifan Wang Yi Shen Michael M. Zavlanos Karl H. Johansson Division of Decision and Control Systems School of Electrical Enginnering and Computer Science KTH Royal Institute of Technology and also with Digital Futures Stockholm Sweden Department of Mechanical Engineering and Materials Science Duke University Durham NC USA

This paper studies a class of strongly monotone games involving non-cooperative agents that optimize their own time-varying cost functions. We assume that the agents can observe other agents' historical actions and choose actions that best respond to other agents' previous actions; we call this a best response scheme. We start by analyzing the convergence rate of this best response scheme for standard time-invariant games. Specifically, we provide a sufficient condition on the strong monotonicity parameter of the time-invariant games under which the proposed best response algorithm achieves exponential convergence to the static Nash equilibrium. We further illustrate that this best response algorithm may oscillate when the proposed sufficient condition fails to hold, which indicates that this condition is tight. Next, we analyze this best response algorithm for time-varying games where the cost functions of each agent change over time. Under similar conditions as for time-invariant games, we show that the proposed best response algorithm stays asymptotically close to the evolving equilibrium. We do so by analyzing both the equilibrium tracking error and the dynamic regret. Numerical experiments on economic market problems are presented to validate our analysis.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Neural Network Architecture Search for Accurate Visual Pose Estimation aboard Nano-UAVs

Deep Neural Network Architecture Search for Accurate Visual ...

引用

IEEE International Conference on Robotics and Automation (ICRA)

作者： E. Cereda L. Crupi M. Risso A. Burrello L. Benini A. Giusti D. Jahier Pagliari D. Palossi Dalle Molle Institute for Artificial Intelligence USI and SUPSI Lugano Switzerland Department of Control and Computer Engineering Politecnico di Torino Turin Italy Department of Electrical Electronic and Information Engineering University of Bologna Bologna Italy Integrated Systems Laboratory ETH Zürich Zürich Switzerland

Miniaturized autonomous unmanned aerial vehicles (UAVs) are an emerging and trending topic. With their form factor as big as the palm of one hand, they can reach spots otherwise inaccessible to bigger robots and safely operate in human surroundings. The simple electronics aboard such robots (sub-100 mW) make them particularly cheap and attractive but pose significant challenges in enabling onboard sophisticated intelligence. In this work, we leverage a novel neural architecture search (NAS) technique to automatically identify several Pareto-optimal convolutional neural networks (CNNs) for a visual pose estimation task. Our work demonstrates how reallife and field-tested robotics applications can concretely leverage NAS technologies to automatically and efficiently optimize CNNs for the specific hardware constraints of small UAVs. We deploy several NAS-optimized CNNs and run them in closed-loop aboard a 27-g Crazyflie nano-UAV equipped with a parallel ultra-low power System-on-Chip. Our results improve the State-of-the-Art by reducing the in-field control error of 32% while achieving a real-time onboard inference-rate of ~10Hz@10mW and ~50Hz@90mW.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust Beamforming Design and Antenna Selection for Dynamic HRIS-aided MISO systems

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Wang, Jintao Zhou, Binggui Ma, Chengzhi Gong, Shiqi Yang, Guanghua Ma, Shaodan University of Macau State Key Laboratory of Internet of Things for Smart City Department of Electrical and Computer Engineering Macau999078 China Beijing Institute of Technology School of Information and Electronics Beijing100081 China Jinan University School of Intelligent Systems Science and Engineering Zhuhai519070 China

In this paper, we propose a dynamic hybrid active-passive reconfigurable intelligent surface (HRIS) to enhance multiple-input-single-output (MISO) communications, leveraging the property of dynamically placing active elements. Specifically, considering the impact of hardware impairments (HWIs), we investigate channel-aware configurations of the receive antennas at the base station (BS) and the active/passive elements at the HRIS to improve transmission reliability. To this end, we propose a penalty-based exact block coordinate descent (PEBCD) algorithm to address the mean-square-error (MSE) minimization problem for the dynamic HRIS-aided MISO system by jointly optimizing the BS receive antenna selection matrix, the reflection phase coefficients, the reflection amplitude matrix, and the mode selection matrix of the HRIS. Numerical simulations demonstrate the significant superiority of our proposed scheme over conventional benchmark schemes. © 1967-2012 IEEE.

关键词： Rectennas

来源：评论

学校读者我要写书评

暂无评论

GINN-LP: A Growing Interpretable Neural Network for Discovering Multivariate Laurent Polynomial Equations

arXiv

引用

arXiv 2023年

作者： Ranasinghe, Nisal Senanayake, Damith Seneviratne, Sachith Premaratne, Malin Halgamuge, Saman AI Optimization and Pattern Recognition Research Group Dept. of Mechanical Eng. University of Melbourne Australia Melbourne School of Design University of Melbourne Australia Department of Electrical and Computer Systems Engineering Monash University Australia

Traditional machine learning is generally treated as a black-box optimization problem and does not typically produce interpretable functions that connect inputs and outputs. However, the ability to discover such interpretable functions is desirable. In this work, we propose GINN-LP, an interpretable neural network to discover the form and coefficients of the underlying equation of a dataset, when the equation is assumed to take the form of a multivariate Laurent Polynomial. This is facilitated by a new type of interpretable neural network block, named the "power-term approximator block", consisting of logarithmic and exponential activation functions. GINN-LP is end-to-end differentiable, making it possible to use backpropagation for training. We propose a neural network growth strategy that will enable finding the suitable number of terms in the Laurent polynomial that represents the data, along with sparsity regularization to promote the discovery of concise equations. To the best of our knowledge, this is the first model that can discover arbitrary multivariate Laurent polynomial terms without any prior information on the order. Our approach is first evaluated on a subset of data used in SRBench, a benchmark for symbolic regression. We first show that GINN-LP outperforms the state-of-the-art symbolic regression methods on datasets generated using 48 real-world equations in the form of multivariate Laurent polynomials. Next, we propose an ensemble method that combines our method with a high-performing symbolic regression method, enabling us to discover non-Laurent polynomial equations. We achieve state-of-the-art results in equation discovery, showing an absolute improvement of 7.1% over the best contender, by applying this ensemble method to 113 datasets within SRBench with known ground-truth equations. © 2023, CC BY-NC-ND.

关键词： Polynomials

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：