检索结果-内蒙古大学图书馆

: the next leap toward native language-based and explainable machine learning framework

Automated Software engineering 2023年第2期30卷 1-28页

作者： Dessureault, Jean-Sébastien Massicotte, Daniel LSSI - Laboratory of Signal and System Integration Electrical and Computer Engineering Department Université du Québec á Trois-Rivières Trois-Rivières Canada

The machine learning frameworks flourished in the last decades, allowing artificial intelligence to get out of academic circles to be applied to enterprise domains. This field has significantly advanced, but there is still some meaningful improvement to reach the subsequent expectations. The proposed framework, named AI $$^{2}$$ , uses a natural language interface that allows non-specialists to benefit from machine learning algorithms without necessarily knowing how to program with a programming language. The primary contribution of the AI $$^{2}$$ framework allows a user to call the machine learning algorithms in English, making its interface usage easier. The second contribution is greenhouse gas (GHG) awareness. It has some strategies to evaluate the GHG generated by the algorithm to be called and to propose alternatives to find a solution without executing the energy-intensive algorithm. Another contribution is a preprocessing module that helps to describe and to load data properly. Using an English text-based chatbot, this module guides the user to define every dataset so that it can be described, normalized, loaded, and divided appropriately. The last contribution of this paper is about explainability. The scientific community has known that machine learning algorithms imply the famous black-box problem for decades. Traditional machine learning methods convert an input into an output without being able to justify this result. The proposed framework explains the algorithm’s process with the proper texts, graphics, and tables. The results, declined in five cases, present usage applications from the user’s English command to the explained output. Ultimately, the AI $$^{2}$$ framework represents the next leap toward native language-based, human-oriented concerns about machine learning framework.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning Based Dynamic Power Control for UAV Mobility Management

arXiv

引用

arXiv 2023年

作者： Meer, Irshad A. Besser, Karl-Ludwig Ozger, Mustafa Poor, H. Vincent Cavdar, Cicek Department of Electrical and Computer Engineering Princeton University United States Division of Communication Systems KTH Royal Institute of Technology Sweden

—Modern communication systems need to fulfill multiple and often conflicting objectives at the same time. In particular, new applications require high reliability while operating at low transmit powers. Moreover, reliability constraints may vary over time depending on the current state of the system. One solution to address this problem is to use joint transmissions from a number of base stations (BSs) to meet the reliability requirements. However, this approach is inefficient when considering the overall total transmit power. In this work, we propose a reinforcement learning-based power allocation scheme for an unmanned aerial vehicle (UAV) communication system with varying communication reliability requirements. In particular, the proposed scheme aims to minimize the total transmit power of all BSs while achieving an outage probability that is less than a tolerated threshold. This threshold varies over time, e.g., when the UAV enters a critical zone with high-reliability requirements. Our results show that the proposed learning scheme uses dynamic power allocation to meet varying reliability requirements, thus effectively conserving energy. Copyright © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Dormant Neural Trojans

Dormant Neural Trojans

引用

International Conference on Machine Learning and Applications (ICMLA)

作者： Feisi Fu Panagiota Kiourti Wenchao Li Division of Systems Engineering Boston University Boston MA USA Department of Electrical and Computer Boston University Boston MA USA

We present a novel methodology for neural network backdoor attacks. Unlike existing training-time attacks where the Trojaned network would respond to the Trojan trigger after training, our approach inserts a Trojan that will remain dormant until it is activated. The activation is realized through a specific perturbation to the network's weight parameters only known to the attacker. Our analysis and the experimental results demonstrate that dormant Trojaned networks can effectively evade detection by state-of-the-art backdoor detection methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Towards a privacy-preserving distributed cloud service for preprocessing very large medical images

arXiv

引用

arXiv 2023年

作者： Wang, Yuandou Kanwal, Neel Engan, Kjersti Rong, Chunming Zhao, Zhiming Multiscale Networked Systems University of Amsterdam Netherlands Department of Electrical Engineering and Computer Science University of Stavanger Norway Amsterdam Netherlands

Digitized histopathology glass slides, known as Whole Slide Images (WSIs), are often several gigapixels large and contain sensitive metadata information, which makes distributed processing unfeasible. Moreover, artifacts in WSIs may result in unreliable predictions when directly applied by Deep Learning (DL) algorithms. Therefore, preprocessing WSIs is beneficial, e.g., eliminating privacy-sensitive information, splitting a gigapixel medical image into tiles, and removing the diagnostically irrelevant areas. This work proposes a cloud service to parallelize the preprocessing pipeline for large medical images. The data and model parallelization will not only boost the end-to-end processing efficiency for histological tasks but also secure the reconstruction of WSI by randomly distributing tiles across processing nodes. Furthermore, the initial steps of the pipeline will be integrated into the Jupyter-based Virtual Research Environment (VRE) to enable image owners to configure and automate the execution process based on resource allocation. © 2023, CC BY-SA.

关键词： Pipelines

来源：评论

学校读者我要写书评

暂无评论

Mini-InternVL: a flexible-transfer pocket multi-modal model with 5% parameters and 90% performance

引用

Visual Intelligence 2024年第1期2卷 1-17页

作者： Gao, Zhangwei Chen, Zhe Cui, Erfei Ren, Yiming Wang, Weiyun Zhu, Jinguo Tian, Hao Ye, Shenglong He, Junjun Zhu, Xizhou Lu, Lewei Lu, Tong Qiao, Yu Dai, Jifeng Wang, Wenhai Shanghai AI Laboratory Shanghai 200232 China School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University Shanghai 200240 China School of Computer Science Nanjing University Nanjing 210023 China Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen 518055 China School of Computer Science Fudan University Shanghai 200433 China SenseTime Research Shanghai 200233 China Department of Electronic Engineering Tsinghua University Beijing 100084 China Beijing National Research Center for Information Science and Technology Department of Automation Tsinghua University Beijing 100084 China Department of Information Engineering The Chinese University of Hong Kong 999077 Hong Kong

Multi-modal large language models (MLLMs) have demonstrated impressive performance in vision-language tasks across a wide range of domains. However, the large model scale and associated high computational cost pose significant challenges for training and deploying MLLMs on consumer-grade GPUs or edge devices, thereby hindering their widespread application. In this work, we introduce Mini-InternVL, a series of MLLMs with parameters ranging from 1 billion to 4 billion, which achieves 90% of the performance with only 5% of the parameters. This significant improvement in efficiency and effectiveness makes our models more accessible and applicable in various real-world scenarios. To further promote the adoption of our models, we are developing a unified adaptation framework for Mini-InternVL, which enables our models to transfer and outperform specialized models in downstream tasks, including autonomous driving, medical image processing, and remote sensing. We believe that our models can provide valuable insights and resources to advance the development of efficient and effective MLLMs. © The Author(s) 2024.

关键词： Knowledge distillation Lightweight multi-modal large language model Vision-language model Visual instruction tuning

来源：评论

学校读者我要写书评

暂无评论

2.4 GHz Ultra-Low Power Direct Digital-to-RF CMOS Transmitter for Biosensing Applications

2.4 GHz Ultra-Low Power Direct Digital-to-RF CMOS Transmitte...

引用

IEEE Biomedical Circuits and systems (BIOCAS)

作者： Fariborz Lohrabi Pour Dong Sam Ha Bradley Department of Electrical and Computer Engineering Multifunctional Integrated Circuits and Systems (MICS) Group Virginia Tech Blacksburg Virginia USA

This paper presents a low-power and low-complexity direct digital-to-RF transmitter architecture, suitable for biosensing applications. The RF front end of the transmitter is based on a ring oscillator, whose output phase is modulated through the charge-to-phase mechanism using a charge injection block. Hence, the phase shift keying (PSK) modulation can be performed directly in the RF domain. Post-layout simulation results show that the transmitter is able to collect, process, and transmit sensed data with the maximum data rate of 20 Mbps and an error vector magnitude (EVM) of smaller than 3.5%, while dissipating the DC power smaller than 0.5 mW. The results demonstrate that the proposed transmitter architecture is effective for wireless biosensing applications.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Guiding AI-Generated Digital Content with Wireless Perception

arXiv

引用

arXiv 2023年

作者： Wang, Jiacheng Du, Hongyang Niyato, Dusit Xiong, Zehui Kang, Jiawen Mao, Shiwen Shen, Xuemin The School of Computer Science and Engineering Nanyang Technological University Singapore The Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore The School of Automation Guangdong University of Technology China The Department of Electrical and Computer Engineering Auburn University Auburn United States The Department of Electrical and Computer Engineering University of Waterloo Canada

Recent advances in artificial intelligence (AI), coupled with a surge in training data, have led to the widespread use of AI for digital content generation, with ChatGPT serving as an representative example. Despite the increased efficiency and diversity, the inherent instability of AI models poses a persistent challenge in guiding these models to produce the desired content for users. In this paper, we introduce an integration of wireless perception (WP) with AI-generated content (AIGC) and propose a unified WP-AIGC framework to improve the quality of digital content production. The framework employs a novel multi-scale perception technology to read user’s posture, which is difficult to describe accurately in words, and transmits it to the AIGC model as skeleton images. Based on these images and user’s service requirements, the AIGC model generates corresponding digital content. Since the production process imposes the user’s posture as a constraint on the AIGC model, it makes the generated content more aligned with the user’s requirements. Additionally, WP-AIGC can also accept user’s feedback, allowing adjustment of computing resources at edge server to improve service quality. Experiments results verify the effectiveness of the WP-AIGC framework, highlighting its potential as a novel approach for guiding AI models in the accurate generation of digital content. © 2023, CC BY.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Optimum sensors allocation for drones multi-target tracking under complex environment using improved prairie dog optimization

引用

Neural Computing and Applications 2024年第18期36卷 10501-10525页

作者： Zitar, Raed Abu Alhadhrami, Esra Abualigah, Laith Barbaresco, Frederic Seghrouchni, Amal ElFallah Thales/SCAI Sorbonne University Abu Dhabi United Arab Emirates Hourani Center for Applied Scientific Research Al-Ahliyya Amman University Amman19328 Jordan Computer Science Department Al al-Bayt University Mafraq25113 Jordan MEU Research Unit Middle East University Amman11831 Jordan Department of Electrical and Computer Engineering Lebanese American University Byblos13-5053 Lebanon Research Center University of Tabuk Tabuk71491 Saudi Arabia Thales Land and Air Systems Paris France Mohammed VI Polytechnic University-Ai movement Rabat Morocco Sorbonne University Paris France College of Engineering Yuan Ze University Taoyuan Taiwan School of Engineering and Technology Sunway University Malaysia Petaling Jaya Malaysia Applied science research center Applied science private university Amman Jordan

This paper presents a novel hybrid optimization method to solve the resource allocation problem for multi-target multi-sensor tracking of drones. This hybrid approach, the Improved Prairie Dog Optimization Algorithm (IPDOA) with the Genetic Algorithm (GA), utilizes the strengths of both algorithms to improve the overall optimization performance. The goal is to select a set of sensors based on norms of weighted distances cost function. The norms are the Euclidean distance and the Mahalanobis distance between the drone location and the sensors. The second one depends on the predicted covariance of the tracker. The Extended Kalman Filter (EKF) is used for state estimation with proper clutter and detection models. Since we use Multi-objects to track, the Joint Probability Distribution Function (JPDA) estimates the best measurement values with a preset gating threshold. The goal is to find a sensor or minimum set of sensors that would be enough to generate high-quality tracking based on optimum resource allocation. In the experimentation simulated with Stone Soup, one radar among five radars is selected at every time step of 50-time steps for 200 tracks distributed over 20 different ground truths. The proposed IPDOA provided optimum solutions for this complex problem. The obtained solution is an optimum offline solution that is used to select one or more sensors for any future flights within the vicinity of the 5 radars. Environment and conditions are assumed to be similar in future drone flights within the radars’ defined zone. The IPDOA performance was compared with the other 8 metaheuristic optimization algorithms and the testing showed its superiority over those techniques for solving this complex problem. The proposed simulated model can find the most relevant sensor(s) capable of generating the best quality tracks based on weighted distance criteria (Euclidean and Mahalanobis). That would cut down the cost of operating extra sensors and then it would be possible to

关键词： Clutter (information theory)

来源：评论

学校读者我要写书评

暂无评论

Nonparametric Sparse Online Learning of the Koopman Operator

arXiv

引用

arXiv 2025年

作者： Hou, Boya Sanjari, Sina Dahlin, Nathan Koppel, Alec Bose, Subhonmesh Carl R. Woese Institute for Genomic Biology University of Illinois Urbana-Champaign United States Department of Mathematics and Computer Science Royal Military College of Canada Canada Department of Electrical and Computer Engineering University at Albany SUNY United States Artificial Intelligence Research JP Morgan Chase & Co Department of Electrical and Computer Engineering Coordinated Science Laboratory University of Illinois Urbana-Champaign United States

The Koopman operator provides a powerful framework for representing the dynamics of general nonlinear dynamical systems. Data-driven techniques to learn the Koopman operator typically assume that the chosen function space is closed under system dynamics. In this paper, we study the Koopman operator via its action on the reproducing kernel Hilbert space (RKHS), and explore the mis-specified scenario where the dynamics may escape the chosen function space. We relate the Koopman operator to the conditional mean embeddings (CME) operator and then present an operator stochastic approximation algorithm to learn the Koopman operator iteratively with control over the complexity of the representation. We provide both asymptotic and finite-time last-iterate guarantees of the online sparse learning algorithm with trajectory-based sampling with an analysis that is substantially more involved than that for finite-dimensional stochastic approximation. Numerical examples confirm the effectiveness of the proposed algorithm. © 2025, CC BY-NC-ND.

关键词： Dynamical systems

来源：评论

学校读者我要写书评

暂无评论

Generative AI-Driven Incentive Mechanism for Semantic Communications in RSMA Networks

引用

IEEE Transactions on Cognitive Communications and Networking 2025年

作者： Liu, Dongqing Liu, Yongwen Zhang, Lei Hafid, Abdelhakim Khoukhi, Lyes Li, Mingyuan Liu, Yutao Dong, Yadong The First Affiliated Hospital of Zhengzhou University National Engineering Laboratory for Internet Medical Systems and Applications Zhengzhou China Zhengzhou University of Light Industry Software Engineering College Zhengzhou China Henan University of Technology College of Electrical Engineering Zhengzhou China University of Montreal Department of Computer Science and Operational Research Canada GREYC Laboratory ENSICAEN Normandie University France

This paper proposes a framework integrating Rate Splitting Multiple Access (RSMA), semantic communications, and generative AI for optimizing next-generation wireless networks. We present a unified model that combines RSMA with semantic communications to enhance both spectral efficiency and semantic fidelity. Our system model focuses on a downlink semantic communication system with a multi-antenna base station serving multiple single-antenna users. A dynamic contract-based incentive mechanism is developed to address user heterogeneity and information asymmetry in semantic RSMA scenarios. We introduce a diffusion model-based approach for joint optimization of resource allocation and contract design in RSMA systems. The semantic encoding process extracts key information, i.e., free-space detection, interest points, object attributes, and spatial relationships, from image data. A loss function is designed to train the semantic encoder and RSMA scheme, incorporating semantic extraction, partitioning, reconstruction, and task specific components. Our framework includes a multi-objective performance evaluation that considers both conventional metrics and semantic accuracy in a multi-user RSMA environment. We also define a multi-component semantic accuracy metric to assess the quality and utility of the extracted semantic information. Extensive simulation results demonstrate the superiority of our proposed framework over existing approaches in terms of system throughput, energy efficiency, and semantic fidelity across various network scenarios and user distributions. © 2015 IEEE.

关键词： Resource allocation

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：