Deep Learning (DL) models are widely used in machine learning due to their performance and ability to deal with large datasets while producing high accuracy and performance metrics. The size of such datasets and the complexity of DL models cause such models to be complex, consuming large amount of resources and time to train. Many recent libraries and applications are introduced to deal with DL complexity and efficiency issues. In this paper, we evaluated one example, Microsoft DeepSpeed library through classification tasks. DeepSpeed public sources reported classification performance metrics on the LeNet architecture. We extended this through evaluating the library on several modern neural network architectures, including convolutional neural networks (CNNs) and Vision Transformer (ViT). Results indicated that DeepSpeed, while can make improvements in some of those cases, it has no or negative impact on others.
Data Quality Toolkit: Automatic assessment of data quality and remediation for machine learning datasets
IBM Research(IBM研究院)
训练数据的质量对机器学习任务的效率、准确性和复杂性有着巨大的影响。有各种工具和技术可用于评估与一般清理和分析检查有关的数据质量。然而,这些技术不适用于检测数据问题的背景下,机器学习任务,如嘈杂的标签,存在重叠的类等,我们试图重新审视数据质量问题的背景下,建立一个机器学习管道,并建立一个工具,可以检测,解释和补救问题的数据,并系统和自动捕获所有的变化应用到数据。我们引入了用于机器学习的数据质量工具包,作为一些关键质量指标和相关补救技术的库,以分析和增强机器学习项目的结构化训练数据集的准备程度。该工具包可以减少数据准备管道的周转时间,并简化数据质量评估过程。我们的工具包通过IBM API Hub [1]平台公开提供,任何开发人员都可以使用IBM的Data Quality for AI API [2]评估数据质量。详细的教程也可以在IBM Learning Path [3]上找到。
Advancing multivariate time series similarity assessment: an integrated computational approach
International Centre of Insect Physiology and Ecology (icipe)(国际昆虫生理与生态中心);Faculty of Science, University of Dschang(大学丁加角学院);University of KwaZulu-Natal(夸祖鲁-纳塔尔大学);African Union Inter-African Phytosanitary Council (AU-IAPSC)(非洲联盟国际非洲植物检疫委员会)
LibEMER: A novel benchmark and algorithms library for EEG-based Multimodal Emotion Recognition
XJTU-POLIMI Joint School, Xi'an Jiaotong University School of Computer Science;Technology, Xi'an Jiaotong University MIGU Video Co., Ltd., Shanghai, China
Metric Hub: A metric library and practical selection workflow for use-case-driven data quality assessment in medical AI
Physikalisch-Technische Bundesanstalt(联邦物理技术研究院);Fraunhofer Institute for Integrated Circuits IIS(弗劳恩霍夫集成电路研究所);Friedrich-Alexander-University Erlangen Nuremberg(弗里德里希-亚历山大-埃尔兰根-纽伦堡大学);Univ Rennes(雷恩大学);Inserm(法国国家健康与医学研究院);Politecnico di Milano(米兰理工大学);Fondazione IRCCS Istituto Nazionale dei Tumori di Milano(米兰国家肿瘤研究所IRCCS基金会);Masaryk University(马萨里克大学);Berlin Institute of Health (BIH)(柏林健康研究所);RISE Research Institutes of Sweden(瑞典RISE研究院);CETIC Centre d’excellence en technologies de l’information et de la communication(CETIC信息与通信技术卓越中心);Fraunhofer Institute for Telecommunications Heinrich-Hertz-Institute HHI(弗劳恩霍夫电信研究所海因里希·赫兹研究所);Multitel Research Center(Multitel研究中心);Technical University Berlin(柏林工业大学);Einstein Centre for Digital Future(爱因斯坦数字未来中心)
Helmholtz-Zentrum Dresden-Rossendorf(德累斯顿-罗森多夫亥姆霍兹中心);Helmholtz Institute Freiberg for Resource Technology(弗赖贝格亥姆霍兹资源技术研究所);Univ. Grenoble Alpes(格勒诺布尔阿尔卑斯大学);Inria(法国国家信息与自动化研究所);CNRS(法国国家科学研究中心);Grenoble INP(格勒诺布尔国立理工学院)
dashi: A Python library for Dataset Shift Characterization to Support Trustworthy AI Development and Deployment
Biomedical Data Science Lab, Instituto Universitario de Tecnologías de la Información y Comunicaciones, Universitat Politècnica de Valéncia(生物医学数据科学实验室,信息与通信技术大学,巴塞罗那理工大学)
Comprehensive Evaluation of Prototype Neural Networks
Semantic Information Systems Group, Osnabrück University, Germany(语义信息系统组,奥斯纳布吕克大学,德国);Research Department Cooperative and Autonomous Systems, German Research Center for Artificial Intelligence (DFKI), Germany(协作与自主系统研究部,德国人工智能研究中心(DFKI),德国)