共查询到19条相似文献，搜索用时 109 毫秒
1.
文本提出了一种基于语义的特征降维方法。通过依存关系抽取实现一次降维;通过计算类别和依存关系特征项的语义相似度,结合互信息方法进行特征选择实现二次降维。对中文文本分类的实验结果表明,提出的特征降维方法具有较好的分类效果。 相似文献
2.
3.
4.
维数简约是肺结节分类识别问题中的关键步骤,现有的方法中都是将所有类别的数据作为一个整体进行降维,忽略了不同类别数据之间在特征子集上的差异性。本文提出了一种将类集和类对相结合的有监督流形特征抽取思想,并将之应用于肺结节的分类中,最终形成一个基于CT影像的肺结节分类系统。实验结果表明了方法的有效性。 相似文献
5.
6.
为了实现对管理领域中主题的准确识别,在从管理活动描述文本中抽取管理问题特定信息的基础上.通过系统考察从文本表示、特征降维到聚类分析的方法技术特点,并提出类特征概化这一新的概念和实现方法.完成了对所研究问题的解决方案设计.实证分析结果与问卷结果和访谈结论所进行的综合分析和判断结果相比较,初步验证了本研究方法对于主题识别的有效性和可行性. 相似文献
7.
特征选择是模式识别中的一个重要问题，本文通过分析特征选择的机理，提出了一种基于层次分析的特征选择法，实现了专家知识和数学方法的有效结合。最后通过数值实验，证明了此方法能较好解决中等规模数据集的降维问题具有较好的结果。 相似文献
8.
癌蛋白质质谱数据中包含了大量未知的内部结构和变量。针对高通量癌蛋白质质谱数据这些特点,分别采用基于PCA-SVM和直接降维+SVM的方法对典型卵巢癌蛋白质质谱数据的进行分类对比实验,实验表明PCA-SVM方法随着PCA降维抽取维数的增加,正确率明显增加,可以较好的对测试样本进行识别和分类。 相似文献
9.
【目的/意义】本文构建一种“特征降维”文本复杂网络进行文本表示,解决传统词同现文本复杂网络处理海 量数据时的维数灾难与语义不足问题,再结合机器学习方法提升文本特征提取效果。【方法/过程】依据共现关系抽 取二级词条,再结合依存句法关系抽取三级词条,构建“特征降维”文本复杂网络,接着利用PCA算法和TOPSIS法 评价网络节点重要性提取反映文本主题的关键词作为文本特征词,实现文本特征提取。【结果/结论】本文以网络新 闻数据为实验对象。实验结果表明,特征降维文本复杂网络能较好地表示中文文本,并且在较好地保留了文本语 义信息的同时有效减少网络节点冗余,结合PCA算法的特征提取方法可以使文本分类性能提高。 相似文献
10.
11.
目标噪声特征提取是被动声纳目标识别系统的关键技术。首先提出了一种利用从噪声极限环中提取的非线性特征来分析舰船噪声信号的新方法,然后采用基于自适应遗传BP算法的神经网络对提取的特征进行分类。实验结果表明,该系统具有较好的分类效果。 相似文献
12.
首先介绍了Hu不变距的相关理论,采用Hu矩对红外图像进行特征提取,并针对提取的Hu矩各分量数值的数量级差别过大,不能同等体现各分量的重要性,以及Hu不变矩对二维离散图像不满足保证缩放不变性的问题,提出了两种改进方法。然后对两种改进方法的尺度收缩性进行比较,采用Matlab实验仿真技术对缩放性能较优的Hu矩改进方法进行了不变性测试并提取特征。 相似文献
13.
14.
本文设计了一种基于RFIC和人脸识别的ARM嵌入式安全认证系统。深入地分析人脸识别过程中的预处理,以及PCA和LDA等特征提取方法,利用MATLAB对人脸识别算法仿真,并比较特征向量个数以及分类器等因素对识别率的影响。 相似文献
15.
《Information processing & management》2023,60(4):103373
Detecting feature interactions is an important post-hoc method to explain black-box models. The literature on feature interactions mainly focus on detecting their existence and calculating their strength. Little attention has been given to the form how the features interact. In this paper, we propose a novel method to capture the form of feature interactions. First, the feature interaction sets in black-box models are detected by the high dimensional model representation-based method. Second, the pairwise separability of the detected feature interactions is determined by a novel model which is verified theoretically. Third, the set separability of the feature interactions is inferred based on pairwise separability. Fourth, the interaction form of each feature in product separable sets is explored. The proposed method not only provides detailed information about the internal structure of black-box models but also improves the performance of linear models by incorporating the appropriate feature interactions. The experimental results show that the accuracy of recognizing product separability in synthetic models is 100%. Experiments on three regression and three classification tasks demonstrate that the proposed method can capture the product separable form of feature interactions effectively and improve the prediction accuracy greatly. 相似文献
16.
Most previous works of feature selection emphasized only the reduction of high dimensionality of the feature space. But in cases where many features are highly redundant with each other, we must utilize other means, for example, more complex dependence models such as Bayesian network classifiers. In this paper, we introduce a new information gain and divergence-based feature selection method for statistical machine learning-based text categorization without relying on more complex dependence models. Our feature selection method strives to reduce redundancy between features while maintaining information gain in selecting appropriate features for text categorization. Empirical results are given on a number of dataset, showing that our feature selection method is more effective than Koller and Sahami’s method [Koller, D., & Sahami, M. (1996). Toward optimal feature selection. In Proceedings of ICML-96, 13th international conference on machine learning], which is one of greedy feature selection methods, and conventional information gain which is commonly used in feature selection for text categorization. Moreover, our feature selection method sometimes produces more improvements of conventional machine learning algorithms over support vector machines which are known to give the best classification accuracy. 相似文献
17.
针对赤潮生物提出具有较高准确率的实时自动分类方法,本文提出采用ReliefF-SBS进行特征选择,即针对赤潮生物图像原始数据集进行特征分析,并在此基础上,对原始特征集进行特征选择以去除特征集中的无关特征和冗余特征,得到最优特征子集,减少它们对分类器分类精度的影响。文中给出了实验结果和分析,同时验证了对k-Nearest Neighbor algorithm(KNN)和Support Vector Machine(SVM)分类器分类效果的影响。 相似文献
18.
特征点匹配是计算机视觉中的关键步骤,在很多领域中都有着的重要应用。通过对当前图像特征点匹配方法的研究,提取一种基于特征点的灰度量和几何特征量相结合的匹配方法。该方法首先用Harris算法提取特征点;然后用极线约束减少搜索范围;最后用特征点的灰度量实现特征点匹配。该方法利用极线约束,克服了用灰度量进行特征点匹配计算量大的缺点,提高了匹配速度。实验表明,是一种准确快速的特征点匹配方法。 相似文献
19.
《Information processing & management》2022,59(1):102767
Think tanks have been proved helpful for decision-making in various communities. However, collecting information manually for think tank construction implies too much time and labor cost as well as inevitable subjectivity. A probable solution is to retrieve webpages of renowned experts and institutes similar to a given example, denoted as query by webpage (QBW). Considering users’ searching behaviors, a novel QBW model based on webpages’ visual and textual features is proposed. Specifically, a visual feature extraction module based on pre-trained neural networks and a heuristic pooling scheme is proposed, which bridges the gap that existing extractors fail to extract snapshots’ high-level features and are sensitive to the noise effect brought by images. Moreover, a textual feature extraction module is proposed to represent textual content in both term and topic grains, while most existing extractors merely focus on the term grain. In addition, a series of similarity metrics are proposed, including a textual similarity metric based on feature bootstrapping to improve model’s robustness and an adaptive weighting scheme to balance the effect of different types of features. The proposed QBW model is evaluated on expert and institute introduction retrieval tasks in academic and medical scenarios, in which the average value of MAP has been improved by 10% compared to existing baselines. Practically, useful insights can be derived from this study for various applications involved with webpage retrieval besides think tank construction. 相似文献