检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Puglisi, Lemuel Rondinella, Alessia De Meo, Linda Guarnera, Francesco Battiato, Sebastiano Ravì, Daniele Dept. of Math and Computer Science University of Catania Catania Italy Department of Neuroinflammation University College London London United Kingdom School of Physics Engineering and CS University of Hertfordshire Hatfield United Kingdom

Brain age is a critical measure that reflects the biological ageing process of the brain. The gap between brain age and chronological age, referred to as brain PAD (Predicted Age Difference), has been utilized to investigate neurodegenerative conditions. Brain age can be predicted using MRIs and machine learning techniques. However, existing methods are often sensitive to acquisition-related variabilities, such as differences in acquisition protocols, scanners, MRI sequences, and resolutions, significantly limiting their application in highly heterogeneous clinical settings. In this study, we introduce Synthetic Brain Age (SynthBA), a robust deep-learning model designed for predicting brain age. SynthBA utilizes an advanced domain randomization technique, ensuring effective operation across a wide array of acquisition-related variabilities. To assess the effectiveness and robustness of SynthBA, we evaluate its predictive capabilities on internal and external datasets, encompassing various MRI sequences and resolutions, and compare it with state-of-the-art techniques. Additionally, we calculate the brain PAD in a large cohort of subjects with Alzheimer’s Disease (AD), demonstrating a significant correlation with AD-related measures of cognitive dysfunction. SynthBA holds the potential to facilitate the broader adoption of brain age prediction in clinical settings, where re-training or fine-tuning is often unfeasible. The SynthBA source code and pre-trained models are publicly available at https://***/LemuelPuglisi/SynthBA. © 2024, CC BY.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automatically Learning HTN Methods from Landmarks

arXiv

引用

arXiv 2024年

作者： Li, Ruoxi Nau, Dana Roberts, Mark Fine-Morris, Morgan Dept. of Computer Science Institute for Systems Research Univ. of Maryland College ParkMD United States Navy Center for Applied Research in AI Naval Research Laboratory WashingtonDC United States

Hierarchical Task Network (HTN) planning usually requires a domain engineer to provide manual input about how to decompose a planning problem. Even HTN-MAKER, a well-known method-learning algorithm, requires a domain engineer to annotate the tasks with information about what to learn. We introduce CURRICULAMA, an HTN method learning algorithm that completely automates the learning process. It uses landmark analysis to compose annotated tasks and leverages curriculum learning to order the learning of methods from simpler to more complex. This eliminates the need for manual input, resolving a core issue with HTN-MAKER. We prove CURRICULAMA’s soundness, and show experimentally that it has a substantially similar convergence rate in learning a complete set of methods to HTN-MAKER. © 2024, CC BY-NC-ND.

关键词： Curricula

来源：评论

学校读者我要写书评

暂无评论

Slotted ALOHA with Code Combining for IoT Networks

Slotted ALOHA with Code Combining for IoT Networks

引用

Communications and Networking (MeditCom), IEEE International Mediterranean Conference on

作者： Dimitrios Tyrovolas Athanasios Chrysologou Georgios Chondrogiannis Sotiris Tegos Prodromos-Vasileios Mekikis Panagiotis Diamantoulakis Sotiris Ioannidis Christos Liaskos Nestor Chatzidiamantis George Karagiannidis Dept. of Electrical and Computer Engineering Aristotle University of Thessaloniki Thessaloniki Greece Dept. of Electrical and Computer Engineering Technical University of Chania Chania Greece Computer Science Engineering Department University of Ioannina Ioannina Greece Cyber Security Systems and Applied AI Research Center Lebanese American University (LAU) Lebanon

Internet of Things (IoT) networks are expected to be a key enabler technology for smart cities, due to their ability to offer real-time monitoring. In this work, we propose a novel medium access control (MAC) protocol, inspired by the synergy of slotted ALOHA and hybrid automatic repeat request with code combining (HARQ-CC), termed as code combining slotted ALOHA (CCSA), to improve the performance of IoT networks with multiple randomly deployed sensors. Specifically, we provide analytical closed-form expressions for the network’s outage probability as well as for its average throughput when CCSA is applied. Finally, we provide simulation results that reveal the authenticity of the derived expressions and reveal the capabilities of HARQ-CC to significantly improve the network’s performance when the sensors’ access is based on a random access scheme.

关键词：

来源：评论

学校读者我要写书评

暂无评论

University’s Educational Environment Personalization Based on the Ontological Models

University’s Educational Environment Personalization Based ...

引用

Annual Siberian Russian Workshop on Electron Devices and Materials (EDM)

作者： Natalya V. Pustovalova Tatiana V. Avdeenko Alexandra V. Pustovalova dept. Automation and Computer Engineering Novosibirsk State Technical University Novosibirsk Russian Federation dept. Applied Mathematics and Computer Science Novosibirsk State Technical University Novosibirsk Russian Federation

ISBN: (数字)9781665498043

ISBN: (纸本)9781665498050

The paper presents a work in progress. Personalization of the educational environment allows improving the quality of the educational process, raising it to a new level. Our development is intended to diagnose and recommend educational content, depending on the students’ personality. This article describes a system of ontological models (Learner Model, educational content model and cognitive functions model) designed to diagnose students’ characteristics and the subsequent issuance of educational content, taking into account these characteristics. The learner model, implemented as a Protege ontology, is based not only on human professional experience but also on student testing results. This approach, based on the use of various IQ tests, allows reflecting in the ontology the dependencies identified due to linear regression methods, as well as other types of analysis. By this way we will reduce the dependence of the model on the human expert’s personality and make it more objective. Using the Protege tools, we will combine three ontologies in the architecture of a personalized environment through rules and axioms.

关键词： Linear regression Ontologies Electron devices Testing

来源：评论

学校读者我要写书评

暂无评论

Exploration and Evaluation of Machine Learning Methods for Detecting Oil Spills in Open Seas

Exploration and Evaluation of Machine Learning Methods for D...

引用

Decision Aid sciences and Application (DASA), International Conference on

作者： Fadi N. Sibai Ahmad Sibai Ali El-Moursy GUST Engineering & Applied Innovation Research Center Gulf Univ. for Science and Tech. West Mishref Kuwait Computer Science & Engg. Dept. Univ. of South Florida Tampa Florida U.S.A. Comp. Engg. Dept. Univ. of Sharjah Sharjah U.A.E.

ISBN: (数字)9798350369106

ISBN: (纸本)9798350369113

Detecting oil spills in open seas is important for the early containment of the pollution diffusion and avoidance of further ecological catastrophes. Herein, we explore and assess the application of several machine learning (ML) methods and image processing techniques for the detection of oil spills in images of open seas captured at the elevation of planes, helicopters, or drones. The image dataset used contains 24 clean sea images and 24 dirty sea images with oil slicks. The machine learning methods experimented with include neural networks, SVM, random forest, decision tree, naïve Bayes, AdaBoost, and logistic regression. We also evaluated these ML methods with and without the application of Sobel Feldman edge detection and Vintage filtering image processing techniques. Results indicate the superiority of the in and neural networks methods over the other explored ML methods, achieving perfect and near-perfect classification, respectively.

关键词： Support vector machines Logistic regression Oils Image edge detection Image processing Helicopters Artificial neural networks Satellite images Random forests Drones

来源：评论

学校读者我要写书评

暂无评论

Fault detection in closed-loop systems using a double residual generator

引用

IFAC-PapersOnLine 2022年第6期55卷 298-303页

作者： Henrik Niemann Niels Kjølstad Poulsen Dept. of Electrical Engineering Technical University of Denmark DK-2800 Kgs. Lyngby Denmark Dept. of Applied Mathematics and Computer Science Technical University of Denmark DK-2800 Kgs. Lyngby Denmark

A double residual generator for fault detection in closed-loop systems is presented in this paper. The residual generator gives the standard residual vector and a dual residual vector as outputs. The standard residual vector can be applied directly for fault detection. The dual residual vector depends on known inputs to the system e.g. reference inputs and auxiliary inputs but decoupled from the disturbance in the system. It is also an input vector that can be applied for active fault detection (AFD). The dual residual vector can be considered as a generalized input vector in connection with AFD in closed-loop systems. This gives the possibility to use all known inputs for detection. The dual residual vector gives a link between passive and AFD in closed-loop systems. The double residual generator is given directly as the left Bezout matrix from the Bezout equation.

关键词： Fault detection residual generators parametric faults closed-loop systems coprime factorization

来源：评论

学校读者我要写书评

暂无评论

Efficient fine-grained road segmentation using superpixel-based CNN and CRF models

arXiv

引用

arXiv 2022年

作者： Zohourian, Farnoush Siegemund, Jan Meuter, Mirko Pauli, Josef Dept. Computer Science and Applied Cognitive Science University of Duisburg-Essen Duisburg Germany Delphi Electronics & Saftey Delphi Wuppertal Germany

Towards a safe and comfortable driving, road scene segmentation is a rudimentary problem in camera-based advanced driver assistance systems (ADAS). Despite of the great achievement of Convolutional Neural Networks (CNN) for semantic segmentation task, the high computational efforts of CNN-based methods is still a challenging area. In recent work, we proposed a novel approach to utilize the advantages of CNN’s for the task of road segmentation at reasonable computational effort. The runtime benefits from using irregular superpixels as basis for the input for the CNN rather than the image grid, which tremendously reduces the input size. Although, this method achieved remarkable low computational time in both training and testing phases, the lower resolution of the superpixel domain yields naturally lower accuracy compared to high cost state of the art methods. In this work, we focus on an refinement of the road segmentation utilizing a Conditional Random Field (CRF). The refinement procedure is limited to the superpixels touching the predicted road boundary to keep the additional computational effort low. Reducing the input to the super-pixel domain allows the CNN’s structure to stay small and efficient to compute while keeping the advantage of convolutional layers and makes them eligible for ADAS. Applying CRF compensates the trade-off between accuracy and computational efficiency. The proposed system obtains comparable performance among the top-performing algorithms on the KITTI road benchmark and it’s fast inference makes it particularly suitable for real-time *** Codes I.2, I.4, I.5, © 2022, CC BY-NC-ND.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Natural Language and LLMs in Human-Robot Interaction: Performance and Challenges in a Simulated Setting

Natural Language and LLMs in Human-Robot Interaction: Perfor...

引用

Human-computer Interaction, Optimization and Robotic Applications (HORA), International Congress on

作者： Kelvin Olaiya Giovanni Delnevo Chiara Ceccarini Chan-Tong Lam Giovanni Pau Paola Salomoni Dept. of Computer Science and Engineering University of Bologna Italy Faculty of Applied Sciences Macao Polytechnic University Macao SAR China Autonomous Robotics Research Center Technology Innovation Institute Abu Dhabi United Arab Emirates

ISBN: (数字)9798331510886

ISBN: (纸本)9798331510893

Natural language provides an intuitive and accessible way for humans to communicate with robots, fostering more natural and flexible interaction across a range of tasks. This study investigates how effectively users can command a robot using natural language within a simulated environment. By employing Gemini Flash 2.0 as the underlying Large Language Model (LLM) to interpret and translate user prompts into executable plans, we explore both the strengths and limitations of this approach. The experiments evaluated user-generated prompts across multiple predefined tasks, revealing a spectrum of outcomes — from successful task completions to errors such as misinterpretations, spatial failures, and hallucinated behaviors where the robot acted on non-existent information. The results highlight how different communication strategies, combining directive and conversational phrasing, influenced task performance. This work contributes to advancing Human-Robot Interaction (HRI) design by emphasizing the potential of LLM-powered systems while addressing the challenges of ambiguity and error resilience in user-driven command structures.

关键词： Human computer interaction Translation Large language models Zero shot learning Human-robot interaction Robots Optimization Resilience

来源：评论

学校读者我要写书评

暂无评论

Key Exchange in the Quantum Era: Evaluating a Hybrid System of Public-Key Cryptography and Physical-Layer Security

Key Exchange in the Quantum Era: Evaluating a Hybrid System ...

引用

Annual Conference on Information sciences and Systems (CISS)

作者： Paul Staat Meik Dörpinghaus Azadeh Sheikholeslami Christof Paar Gerhard Fcllwcis Dennis Goeckel Max Planck Institute for Security and Privacy Bochum Germany Vodafone Chair Mobile Communications Systems Technische Universität Dresden Dresden Germany Math and Computer Science Department Suffolk University Boston Dept. Electrical and Computer Engineering University of Massachusetts Amherst USA

ISBN: (数字)9798331513269

ISBN: (纸本)9798331513276

Today's information society relies on cryptography to achieve security goals such as confidentiality, integrity, authentication, and non-repudiation for digital communications. Here, public-key cryptosystems play a pivotal role to share encryption keys and create digital signatures. However, quantum computers threaten the security of traditional public-key cryptosystems as they can tame computational problems underlying the schemes, i.e., discrete logarithm and integer factorization. The prospective arrival of capable-enough quantum computers already threatens today's secret communication in terms of their long-term secrecy when stored to be later decrypted. Therefore, researchers strive to develop and deploy alternative *** this work, we evaluate a key exchange protocol based on combining public-key schemes with physical-layer security, anticipating the prospect of quantum attacks: If a powerful quantum attacker cannot immediately obtain a private key, legitimate parties have a window of short-term secrecy to perform a physical-layer jamming key exchange (JKE) to establish a long-term shared secret. Thereby, the protocol constraints the computation time available to the attacker to break the employed public-key cryptography. In this paper, we outline the protocol, discuss its security, and point out challenges to be resolved.

关键词： computers Protocols Quantum computing Quantization (signal) Non-repudiation Public key cryptography Security Time factors Jamming Signal resolution

来源：评论

学校读者我要写书评

暂无评论

Hyperparameter Learning of Bayesian Context Tree Models

Hyperparameter Learning of Bayesian Context Tree Models

引用

IEEE International Symposium on Information Theory

作者： Yuta Nakahara Shota Saito Koshi Shimada Toshiyasu Matsushima Center for Data Science Waseda University Tokyo Japan Faculty of Informatics Gunma University Gunma Japan Dept. of Pure and Applied Math. Waseda University Tokyo Japan

In recent years, Bayesian counterparts of the context tree weighting method are studied for many tasks. All these tasks require a hyperparameter setting of the prior distribution for context tree models. Therefore, we provide a framework for statistically learning these hyperparameters from data. Specifically, we consider a hierarchical Bayesian model that assumes hyperprior distributions behind the hyperparameters and learn them using an empirical variational Bayesian (EVB) method. This is the first study to propose an EVB method on the Bayesian context trees. The derived algorithm has a suggestive form that consists of subroutines partially optimal to each local probabilistic model.

关键词：

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：