检索结果-内蒙古大学图书馆

A Supervised Constructive Neuro-Immune Network for Pattern Classification

学校读者我要写书评

暂无评论

A Supervised Constructive Neuro-Immune Network for Pattern C...

International Joint Conference on Neural Networks (IJCNN)

作者： H. Knidel L. Nunes de Castro F.J. Von Zuben Laboratory of Bioinformatics and Bioinspired Computing-LBiC Department of Computer Engineering and Industrial Automation-DCA School of Electrical and Computer Engineering-FEEC State University of Campinas-UNICAMP Sao Paulo Brazil Graduate Program in Computer Science Catholic University of Santos Sao Paulo Brazil

This paper proposes a supervised version of a learning algorithm for a constructive neuro-immune network. The proposed methodology is developed by taking ideas from the immune system and learning vector quantization. The resulting classification algorithm is characterized by high-performance, similar to the ones produced by alternative methods in the literature, and parsimonious solutions, with a much smaller set of prototypes per class when compared with the other approaches. The number of prototypes is automatically defined by the convergence criterion. The algorithm requires a single user-defined parameter for training, associated with the convergence criterion, and the computational cost is sufficiently reduced to support applications involving large data sets.

关键词： Pattern classification Immune system Neurons Supervised learning Vector quantization Classification algorithms Prototypes Clustering algorithms Unsupervised learning Artificial neural networks

Fusion of Mixture of Experts and Generative Artificial Intelligence in Mobile Edge Metaverse

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Guangyuan Du, Hongyang Niyato, Dusit Kang, Jiawen Xiong, Zehui Jamalipour, Abbas Mao, Shiwen Kim, Dong In The College of Computing and Data Science The Energy Research Institute @ NTU Interdisciplinary Graduate Program Nanyang Technological University Singapore The College of Computing and Data Science Nanyang Technological University Singapore The School of Automation Guangdong University of Technology China The Pillar of Information Systems Technology and Design Singapore University of Technology and Design Singapore The Department of Electrical and Computer Engineering Auburn University Auburn United States The School of Electrical and Computer Engineering University of Sydney Australia The Department of Electrical and Computer Engineering Sungkyunkwan University Korea Republic of

In the digital transformation era, Metaverse offers a fusion of virtual reality (VR), augmented reality (AR), and web technologies to create immersive digital experiences. However, the evolution of the Metaverse is slowed down by the challenges of content creation, scalability, and dynamic user interaction. Our study investigates an integration of Mixture of Experts (MoE) models with Generative Artificial Intelligence (GAI) for mobile edge computing to revolutionize content creation and interaction in the Metaverse. Specifically, we harness an MoE model's ability to efficiently manage complex data and complex tasks by dynamically selecting the most relevant experts running various sub-models to enhance the capabilities of GAI. We then present a novel framework that improves video content generation quality and consistency, and demonstrate its application through case studies. Our findings underscore the efficacy of MoE and GAI integration to redefine virtual experiences by offering a scalable, efficient pathway to harvest the Metaverse's full potential. © 2024, CC BY.

关键词： Augmented reality

Is Attention All You Need for EEG to Predict Neurological Outcomes in Cardiac Arrest?

学校读者我要写书评

暂无评论

Is Attention All You Need for EEG to Predict Neurological Ou...

2024 IEEE-EMBS International Conference on Biomedical and Health Informatics, BHI 2024

作者： Dionisio, Jefferson Rizwan, Muhammad Suzuki, Hidetaka Hassan, Kazi Mahmudul Chanpornpakdi, Ingon Lin, Lian-Yu Tanaka, Toshihisa Lin, Che Tokyo University of Agriculture and Technology Tokyo Japan Japanese Red Cross Musashino Hospital Department of Emergency and Critical Care Tokyo Japan National Taiwan University College of Medicine and Hospital Division of Cardiology Department of Internal Medicine Taipei Taiwan Artificial Intelligence of Things Taiwan International Graduate Program Academia Sinica Taipei Taiwan National Taiwan University Department of Electrical Engineering Taipei Taiwan Graduate Institute of Communication Engineering National Taiwan University Taipei Taiwan National Taiwan University Center for Advanced Computing and Imaging in Biomedicine Taipei Taiwan Department of Smart Medicine and Health Informatics National Taiwan University Taipei Taiwan

ISBN: (纸本)9798350351552

Sudden cardiac arrest (SCA) poses a significant health challenge, necessitating accurate predictions of neurological outcomes in comatose patients, where good outcomes are defined as the recovery of most cognitive functions. Electroencephalogram (EEG) serves as a valuable biomarker for monitoring neurological states due to its rich, time-dependent information. This study aims to predict neurological outcomes using early EEG data by employing the Transformer model, which leverages multi-headed attention to identify patterns in lengthy sequences such as hour-long EEG recordings. Unlike traditional methods that use subsampled EEG epochs, we utilize the entire EEG sequences, subdivided into time steps, allowing our model to capture detailed temporal patterns via the attention mechanism. Moreover, we trained our proposed model using each EEG recording as an individual data sample but evaluated our model through aggregated patient-wise predictions. This approach allows us to boost the data sample size. Our results demonstrate promising predictive performance, achieving an AUROC of 0.82 and AUPRC of 0.90 on the holdout test set and an AUROC of 0.73 and AUPRC of 0.93 on an external test set with patient-wise predictions. This study highlights the potential of utilizing attention mechanisms to capture important time series progressions across EEG sequences for improving SCA prognosis. © 2024 IEEE.

关键词： Electroencephalography

Valley notch filter in a graphene strain superlattice: a Green’s function and machine learning approach

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Torres, V. Silva, P. De, E.A.T. Souza Silva, L.A. Bahamon, D.A. Instituto de Física Universidade Federal Fluminense Niterói Av. Litorânea sn RJ24210-340 Brazil MackGraphe Mackenzie Presbyterian University Rua da Consolação 930 SP01302-907 Brazil School of Computing and Informatics Graduate Program in Electrical Engineering and Computing

The valley transport properties of a superlattice of out-of-plane Gaussians deformations are calculated using a Green’s function and a Machine Learning approach. Our results show that periodicity significantly improves the valley filter capabilities of a single Gaussian deformation, these manifest themselves in the conductance as a sequence by valley filter plateaus. We establish that the physical effect behind the observed valley notch filter is the coupling between counter-propagating transverse modes;the complex relationship between the design parameters of the superlattice and the valley filter effect make difficult to estimate in advance the valley filter potentialities of a given superlattice. With this in mind, we show that a Deep Neural Network can be trained to predict valley transmission with a precision similar to the Green’s function but with much less computational effort. Copyright © 2019, The Authors. All rights reserved.

关键词： Landforms

Self-supervised approach for facial movement based optical flow

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Alkaddour, Muhannad Tariq, Usman Dhall, Abhinav Graduate Program in Mechatronics Engineering American University of Sharjah Sharjah United Arab Emirates Department of Electrical Engineering American University of Sharjah Sharjah United Arab Emirates Department of Human Centred Computing Monash University Melbourne Australia Department of Computer Science and Engineering Indian Institute of Technology Ropar Rupnagar India

computing optical flow is a fundamental problem in computer vision. However, deep learning-based optical flow techniques do not perform well for non-rigid movements such as those found in faces, primarily due to lack of the training data representing the fine facial motion. We hypothesize that learning optical flow on face motion data will improve the quality of predicted flow on faces. The aim of this work is threefold: (1) exploring self-supervised techniques to generate optical flow ground truth for face images;(2) computing baseline results on the effects of using face data to train Convolutional Neural Networks (CNN) for predicting optical flow;and (3) using the learned optical flow in micro-expression recognition to demonstrate its effectiveness. We generate optical flow ground truth using facial key-points in the BP4D-Spontaneous dataset. The generated optical flow is used to train the FlowNetS architecture to test its performance on the generated dataset. The performance of FlowNetS trained on face images surpassed that of other optical flow CNN architectures, demonstrating its usefulness. Our optical flow features are further compared with other methods using the STSTNet micro-expression classifier, and the results indicate that the optical flow obtained using this work has promising applications in facial expression analysis. © 2021, CC BY.

关键词： Optical flows

Analysis of Negative Capacitance MOSFETs Characteristic with Spacer

学校读者我要写书评

暂无评论

Analysis of Negative Capacitance MOSFETs Characteristic with...

International Symposium on VLSI Technology, Systems and Applications

作者： Yu-Chun Lee Sekhar Reddy Kola Chieh-Yang Chen Min-Hui Chuang Yiming Li Parallel and Scientific Computing Laboratory Institute of Biomedical Engineering EECS International Graduate Program Institute of Communications Engineering Department of Electrical and Computer Engineering National Chiao Tung University Hsinchu Taiwan Center for mmWave Smart Radar System and Technologies National Chiao Tung University 1001 Ta-Hsueh Rd Hsinchu Taiwan

ISBN: (数字)9781728142326

ISBN: (纸本)9781728142333

In this work, our study comprises of design and investigation on negative capacitance (NC), metal-oxide-semiconductor (MOS) field effects transistors (MOSFETs) with spacer and source/drain (S/D) overlap engineering. The scope of the work is to boost the performance and high-energy efficiency of the studied NC-MOSFETs by using the ferro electric material (FE). The NC-MOSFETs with the spacer technology can achieve the admirable I on /I off ratio and subthreshold swing (SS), compared with planar MOSFETs. It makes device scaling possible by eliminating the short channel effect (SCE). We further estimated the effect of FE thickness and spacer, which are another critical parameter of obtaining better electrical characteristics and reducing SS.

关键词： Capacitance Iron Semiconductor device modeling MOSFET Integrated circuit modeling Mathematical model Ferroelectric materials

An Evolutionary Algorithm for Quadcopter Trajectory Optimization in Aerial Challenges

学校读者我要写书评

暂无评论

An Evolutionary Algorithm for Quadcopter Trajectory Optimiza...

IEEE Latin American Robotics Symposium, LARS

作者： Adson Nogueira Alves Murillo Augusto S. Ferreira Esther Luna Colombini Alexandre da Silva Simões Graduate program in Electrical Engineering (PGEE) at the Sao Paulo State University (Unesp) Sorocaba SP Brazil Institute of Computing (IC) of the University of Campinas (Unicamp) Sorocaba SP Brazil Institute of Science and Technology (ICT) of the Sao Paulo State University (Unesp) Sorocaba SP Brazil

ISBN: (数字)9780738111537

ISBN: (纸本)9781665419857

Machine learning methods have been widely employed in robotics over the years, and recent developments in machine learning have completely re-shaped problem-solving in the area. Indeed, if we consider multi-objective planning, these models' optimization and learning capabilities can derive more robust strategies. Inspired by the species natural selection mechanism, Evolutionary Algorithms (EA) are among the best known computational approaches available for this purpose. In this scenario, this work proposed an EA model developed to find the best travel trajectory for a quadcopter in the “Desafio Petrobras” challenge. In the challenge, a set of landing platforms that the robot has to visit are displaced in the 3D-space. To find the best trajectory possible, we optimize an EA over a low-level control that can take the quadcopter from point A to B. We vary our fitness function to support more complex decisions. The software-in-the-loop technique was applied for a simulated quadrotor in the Coppelia simulated environment. The proposed approach has shown the capability to generate short trajectories while considering variables like UAV dynamics and energy consumption.

关键词： Trajectory Statistics Sociology Genetics Genetic algorithms Vehicle dynamics Unmanned aerial vehicles

A bee-inspired algorithm for optimal data clustering

学校读者我要写书评

暂无评论

A bee-inspired algorithm for optimal data clustering

Congress on Evolutionary Computation

作者： Dávila Patrícia Ferreira Cruz Renato Dourado Maia Alexandre Szabo Leandro Nunes de Castro Educational Foundation of Montes Claros Montes Claros Minas Gerais Brazil Computer Sciences Department State University of Montes Claros Montes Claros Minas Gerais Brazil Natural Computing Laboratory (LCoN) Graduate Program in Electrical Engineering Mackenzie Presbyterian University Sao Paulo Brazil

ISBN: (纸本)9781479904532

The amount of data generated in different knowledge areas has made it necessary the use of data mining tools capable of automatically analyzing and extracting knowledge from datasets. Clustering is one of the most important tasks in data mining and can be defined as the process of partitioning objects into groups or clusters, such that objects in the same group are more similar to one another than to objects belonging to other groups. In this context, this paper aims to propose an adaptation of a bee-inspired optimization algorithm so that it is able to solve data clustering problems. The algorithm was run for different datasets and the results obtained showed high quality clusters and diversity of solutions, whilst a suitable number of clusters was automatically determined.

关键词： Clustering algorithms Entropy Algorithm design and analysis Optimization Data mining Computers Sociology

Supervised Score-Based Modeling by Gradient Boosting

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhao, Changyuan Du, Hongyang Liu, Guangyuan Niyato, Dusit College of Computing and Data Science Nanyang Technological University Singapore CNRS@CREATE 1 Create Way # 08-01 Create Tower 138602 Singapore Department of Electrical and Electronic Engineering University of Hong Kong Hong Kong The Energy Research Institute @ NTU Interdisciplinary Graduate Program Singapore

Score-based generative models can effectively learn the distribution of data by estimating the gradient of the distribution. Due to the multi-step denoising characteristic, researchers have recently considered combining score-based generative models with the gradient boosting algorithm, a multi-step supervised learning algorithm, to solve supervised learning tasks. However, existing generative model algorithms are often limited by the stochastic nature of the models and the long inference time, impacting prediction performances. Therefore, we propose a Supervised Score-based Model (SSM), which can be viewed as a gradient boosting algorithm combining score matching. We provide a theoretical analysis of learning and sampling for SSM to balance inference time and prediction accuracy. Via the ablation experiment in selected examples, we demonstrate the outstanding performances of the proposed techniques. Additionally, we compare our model with other probabilistic models, including Natural Gradient Boosting (NGboost), Classification and Regression Diffusion Models (CARD), Diffusion Boosted Trees (DBT), and non-probabilistic GBM models. The experimental results show that our model outperforms existing models in both accuracy and inference time. © 2024, CC BY.

关键词： Self-supervised learning