共查询到20条相似文献,搜索用时 46 毫秒
1.
在很多真实世界问题中,不同类别的数据样本往往有显著的不平衡性,即大类的样本远多于小类.对类别不平衡样本进行学习,是目前国内外数据挖掘和机器学习领域的研究热点之一.以往对不平衡样本学习的研究主要针对二分类问题进行,由此针对多分类问题,提出一种基于HDDT决策树集成的多类不平衡学习方法.实验表明,该方法可以有效地对多类不平衡问题进行学习. 相似文献
2.
少样本学习是目前机器学习研究领域的热点和难点.针对现有的少样本学习模型不能有效捕捉数据特征与数据标签之间的联系,造成分类模型泛化能力弱的问题,提出一种基于元学习的原型空间图卷积网络少样本学习模型FSL-GCNPS(Few-Shot Learning of Graph Convolutional Network on Prototype Space).首先,利用卷积神经网络提取多任务数据的特征向量;其次,为了将特征向量映射到原型空间中,根据元学习的训练策略得到特征向量的类原型表达;然后,通过类原型向量和类向量之间的嵌入表示,构建图结构数据,并进行图卷积网络训练、推理.实验结果表明,相较于经典少样本学习方法,FSL-GCNPS模型拥有更好的分类准确率和分类稳定性.同时,在医学图像领域数据集上实验表明,FSL-GCNPS具有很好的跨域适应性. 相似文献
3.
4.
5.
新一代人工智能技术的特征,表现为借助GPU计算、云计算等高性能分布式计算能力,使用以深度学习算法为代表的机器学习算法,在大数据上进行学习训练,来模拟、延伸和扩展人的智能。不同数据来源、不同的计算物理位置,使得目前的机器学习面临严重的隐私泄露问题,因此隐私保护机器学习(PPM)成为目前广受关注的研究领域。采用密码学工具来解决机器学习中的隐私问题,是隐私保护机器学习重要的技术。该文介绍隐私保护机器学习中常用的密码学工具,包括通用安全多方计算(SMPC)、隐私保护集合运算、同态加密(HE)等,以及应用它们来解决机器学习中数据整理、模型训练、模型测试、数据预测等各个阶段中存在的隐私保护问题的研究方法与研究现状。 相似文献
6.
7.
图像分类作为计算机视觉领域中的重要研究方向之一,应用领域非常广泛.基于深度学习的图像分类技术取得的成功,依赖大量的已标注数据,然而数据的标注成本往往是昂贵的.主动学习作为一种机器学习方法,旨在以尽可能少的高质量标注数据达到期望的模型性能,缓解监督学习任务中存在的标注成本高、标注信息难以大量获取的问题.主动学习图像分类算法根据样本选择策略,从未标记样本数据集合中选择出信息量丰富,对分类模型训练贡献更高的样本进行标注,以更新已标注训练数据池,如此循环直至满足给定的停止条件或模型标注预算耗尽.本文对近年来提出的主动学习图像分类算法进行了详细综述,并根据所用样本数据处理及模型优化方案,将现有算法分为三类:基于数据增强的算法,包括利用图像增广来扩充训练数据,或者根据图像特征插值后的差异性来选择高质量的训练数据;基于数据分布信息的算法,根据数据分布的特点来优化样本选择策略;优化模型预测的算法,包括优化获取和利用深度模型预测信息的方法、基于生成对抗网络和强化学习来优化预测模型的结构,以及基于Transformer结构提升模型预测性能,以确保模型预测结果的可靠性.此外,本文还对各类主动学习图像分类算法... 相似文献
8.
针对磁盘数据集中正负样本数目严重不平衡导致基于机器学习的分类算法易出现故障预测准确率低的问题,本文提出一种基于自适应加权Bagging-GBDT算法的磁盘故障预测模型.首先,提出基于聚类的分层欠采样方法对健康磁盘样本进行多次抽样,解决随机欠采样方法易丢弃潜在有用样本的问题;其次,将每次采样后样本与全部故障磁盘样本组合得到多个样本子集,通过训练这些子集建立多个预测精度较高的GBDT子分类模型;最后,根据待测点邻域样本类别自适应确定各子模型权重,据此通过加权硬投票集成最终的磁盘故障预测模型.在8组KEEL不平衡数据集上实验结果表明,与现有典型不平衡学习算法相比,少数类的召回率平均提升了9.46%;同时在磁盘公开数据集和某调度系统磁盘数据上对比验证了该方法在故障预测率上的先进性. 相似文献
9.
10.
11.
Deep neural network models with strong feature extraction capacity are prone to overfitting and fail to adapt quickly to new tasks with few samples. Gradient-based meta-learning approaches can minimize overfitting and adapt to new tasks fast, but they frequently use shallow neural networks with limited feature extraction capacity. We present a simple and effective approach called Meta-Transfer-Adjustment learning (MTA) in this paper, which enables deep neural networks with powerful feature extraction capabilities to be applied to few-shot scenarios while avoiding overfitting and gaining the capacity for quickly adapting to new tasks via training on numerous tasks. Our presented approach is classified into two major parts, the Feature Adjustment (FA) module, and the Task Adjustment (TA) module. The feature adjustment module (FA) helps the model to make better use of the deep network to improve feature extraction, while the task adjustment module (TA) is utilized for further improve the model’s fast response and generalization capabilities. The proposed model delivers good classification results on the benchmark small sample datasets MiniImageNet and Fewshot-CIFAR100, as proved experimentally. 相似文献
12.
The letter reports on a previously unseen similarity between an associative memory model recently proposed for the brain and some electronic systems that have been developed for practical learning machines. This similarity is analysed, and its implications are discussed. Results of a comparison between the learning behaviour of an electronic network and a group of human subjects in a pattern-recognition task are given to complete the comparative study. 相似文献
13.
Organizational learning (OL) has been considered to be an effective developmental initiative in organizations that have a mission to develop, to sustain and to advance. There is a wide range of beliefs of thinking about what organizational learning is, how it occurs, and how it is applied and how it influences organizational development. Different perspectives are used to study OL by researchers from different disciplines. There is no clear and widely accepted depiction of the linkage between OL and its enabling factors. Such a comprehensive concept covering so many aspects is difficult to achieve. The authors introduce a variant of a classic approach to analyze the OL process?the DELO (Driving-Enabling-Learning-Outcome) model, by describing OL as a continuous process. The model not only describes OL as a continuous goal-driven process, but also allows the picturing of the relationships with various influencing factors and facilitating effect of the of an OL framework. 相似文献
14.
虚拟学社是计算机技术与多媒体网络技术的产物,是一种网上学习环境.从虚拟学社和协作式学习的定义出发,探讨了基于虚拟学社的协作式学习模式的内涵,提出该学习模式的设计原则,详细阐述了虚拟学社协作学习平台的建构,以及该学习模式的实现方式. 相似文献
15.
Chang Wook Kim Seongwon Cho Choong Woong Lee 《Signal Processing: Image Communication》1995,6(6):499-505
In this paper, we present a new competitive learning algorithm with classified learning rates, and apply it to vector quantization of images. The basic idea is to assign a distinct learning rate to each reference vector. Each reference vector is updated independently of all the other reference vectors using its own learning rate. Each learning rate is changed only when its corresponding reference vector wins the competition, and the learning rates of the losing reference vectors are not changed. The experimental results obtained with image vector quantization show that the proposed method learns more rapidly and yields better quality of the coded images than conventional competitive learning method with a scalar learning rate. 相似文献
16.
Today's interactive installations are characterized by simplistic feedback mechanisms, limited underlying databases of content, and very little autonomous intelligence. Yet even these rudimentary systems provide great opportunities for curious learning, surprise, and serendipitous discovery. They also serve as a social nexus. The interpersonal, back-channel communications and ancillary activities of the audience, which currently remain largely unsensed and unprocessed, can be just as important as the primary authored experience. The key to accessing and communicating larger perceptions in the digital age may well lie in the collective social activities that occur within the context of shared narratives. The exchanges and experiences of group exploration and discovery, communal curious learning promises rewards far beyond the mere enjoyment of story and the instant gratifications of an individual's remote control 相似文献
17.
虚拟学习社区是一个通过信息通讯技术(ICT)在互联网或局域网内创建的学习环境。它为学生创造了广阔的协作学习空间,冲击着传统的教学方式,改变着教育的模式。通过定量研究,以数据结构实验课为载体,对学生在基于Groove的虚拟学习社区中进行协作学习的态度、学习成绩、信息素养和网络互动能力的分析,得出了具有一定实用价值的策略。 相似文献
18.