Search - feature

Search - feature - List

[DataMining] pujulei DL : 0: 谱聚类算法建立在谱图理论基础上，与传统的聚类算法相比，它具有能在任意形状的样本空间上聚类且收敛于全局最优解的优点。该算法首先根据给定的样本数据集定义一个描述成对数据点相似度的亲合矩阵,并且计算矩阵的特征值和特征向量，然后选择合适的特征向量聚类不同的数据点。谱聚类算法最初用于计算机视觉、VLS I 设计等领域，最近才开始用于机器学习中，并迅速成为国际上机器学习领域的研究热点。-Spectral clustering algorithm based on the spectrum based on the theory, as compared with traditional clustering algorithm, which has the advantage of clustering and convergence in the sample space to the global optimal solution of arbitrary shape. Firstly, according to the given sample data set defines a description paired data points of similarity affinity matrix, and compute eigenvalues and eigenvectors, and then the appropriate feature vector clustering of different data points. Spectral clustering algorithm was originally used in computer vision, VLS I design and other fields, only recently began to be used in machine learning, and quickly became the focus of international research on the field of machine learning.
Date : 2026-01-10 Size : 1kb User : qiong hiong
[DataMining] R-Code DL : 0: This contains R code for feature selection
Date : 2026-01-10 Size : 2kb User : Arunkumar C
[DataMining] DeepLearning-master DL : 0: 深度学习的概念源于人工神经网络的研究。含多隐层的多层感知器就是一种深度学习结构。深度学习通过组合低层特征形成更加抽象的高层表示属性类别或特征，以发现数据的分布式特征表示。[1] 深度学习的概念由Hinton等人于2006年提出。基于深信度网(DBN)提出非监督贪心逐层训练算法，为解决深层结构相关的优化难题带来希望，随后提出多层自动编码器深层结构。此外Lecun等人提出的卷积神经网络是第一个真正多层结构学习算法，它利用空间相对关系减少参数数目以提高训练性能。[1] 深度学习是机器学习研究中的一个新的领域，其动机在于建立、模拟人脑进行分析学习的神经网络，它模仿人脑的机制来解释数据，例如图像，声音和文本。同机器学习方法一样，深度机器学习方法也有监督学习与无监督学习之分．不同的学习框架下建立的学习模型很是不同．例如，卷积神经网络（Convolutional neural networks，简称CNNs）就是一种深度的监督学习下的机器学习模型，而深度置信网（Deep Belief Nets，简称DBNs）就是一种无监督学习下的机器学习模型。 -Research on the concept of deep learning the artificial neural network. Multi hidden layer of multi-layer perceptron is a deep learning structure. Deep learning features a more abstract representation of a higher level representation of a feature class or feature to find data. [1] Deep learning s concept by Hinton et al. In 2006. Based on the (DBN), a non supervised greedy layer by layer training algorithm is proposed to solve the problems of the deep structure. In addition, the convolutional neural network proposed by Lecun et al is the first real multi layer structure learning algorithm, which uses the relative relationship between the number of parameters to improve the training performance. [1] Deep learning is a new field in machine learning research. The motivation is to establish and simulate the human brain to analyze the learning of neural network, which simulates the human brain mechanism to explain the data, such as image, sound and text. Convolutional (neural) net
Date : 2026-01-10 Size : 66kb User : Francis
[DataMining] Ga_feature_selector_mcode DL : 0: Genetic algorithm for feature selection
Date : 2026-01-10 Size : 2kb User : MANDHATYA SINGH
[DataMining] InfoRetri DL : 0: 基于朴素贝叶斯的文本分类，包含去停用词，分词，特征提取，分类等-Text classification, based libsvm, included to stop words, segmentation, feature extraction and classification
Date : 2026-01-10 Size : 8mb User : xuehui
[DataMining] 1 DL : 0: 检测中文文章的相似度，首先对中文文章分词处理，然后提取特征，计算特征向量夹角。检验是否相似-Similarity detection Chinese article, the first article of the Chinese word processing and feature extraction, feature vector angle calculation. Test whether similar
Date : 2026-01-10 Size : 361kb User : ww
[DataMining] neighborhood-mu-info DL : 0: 基于邻里互信息特征的评价与选择,数据挖掘，Matlab平台-neighborhood mutual information based feature uation and selection
Date : 2026-01-10 Size : 2kb User : liss
[DataMining] wrappers DL : 0: 文档中是有关于数据挖掘中的wrapper特征选择法（sequential feature selection）的详细资料。-The document is about the wrapper feature selection method (aka sequential feature selection).
Date : 2026-01-10 Size : 3.05mb User : Howard Lee
[DataMining] FCBF DL : 0: This a one of the data mining algorithm completly made by me. This is based on feature subset selection. One of the popular algorithm of feature selection named Fast Correlation based feature selection algorithm(FCBF).It s complete code is here with other function files used in it. I can provide you with more description of it if needed.-This is a one of the data mining algorithm completly made by me. This is based on feature subset selection. One of the popular algorithm of feature selection named Fast Correlation based feature selection algorithm(FCBF).It s complete code is here with other function files used in it. I can provide you with more description of it if needed.
Date : 2026-01-10 Size : 2kb User : Arpita
[DataMining] CHI DL : 0: 计算特征的卡方校验值，最后一列为样本标签，输出为特征的卡卡校验值及对应的特征序列号-The chi-square calculate checksum feature, the last one as a sample label, the output characteristics of the card verification value and the corresponding serial number feature
Date : 2026-01-10 Size : 1kb User : xuyan
[DataMining] mixBern DL : 0: Just like EM of Gaussian Mixture Model, this is the EM algorithm for fitting Bernoulli Mixture Model. GMM is useful for clustering real value data. However, for binary data (such as bag of word feature) Bernoulli Mixture is more suitable.
Date : 2026-01-10 Size : 3kb User : lin
[DataMining] pca DL : 0: this code is a feature extraction method function.
Date : 2026-01-10 Size : 1kb User : rozy
[DataMining] svd DL : 0: 奇异值分解在某些方面与对称矩阵或厄米矩阵基于特征向量的对角化类似。然而这两种矩阵分解尽管有其相关性，但还是有明显的不同。-Singular value decomposition in some respects symmetric matrix or Hermitian matrix based on a similar feature vectors diagonalization. However, the two matrix decomposition in spite of its relevance, but there are significant differences.
Date : 2026-01-10 Size : 1kb User : 李琴
[DataMining] Relief DL : 0: Relief算法是一种特征权重算法,可以用于特征选择-Relief algorithm is a feature weighting algorithm，which can be used for feature selection
Date : 2026-01-10 Size : 769kb User : 谈文艺
[DataMining] CARS_Characteristic-band-extraction DL : 0: 特征带提取算法，可以很好地提取有用的频带，结果更为准确，CARS_Characteristic带提取.m为主程序，调用程序中的压缩包。-Feature band extraction algorithm, can be a good extraction of useful bands, the results are more accurate, CARS_Characteristic band extraction.m for the main program, call the program in the compressed package
Date : 2026-01-10 Size : 392kb User : 刘
[DataMining] gini DL : 0: Gini index Calculator. useful in feature selection
Date : 2026-01-10 Size : 1kb User : Mo3000
[DataMining] lasso DL : 0: 使用lasso方法，对特征矩阵进行优化特征选择（向量选择），使其达到最优。-Use lasso method, optimize the characteristic matrix feature selection (vector selection), to the optimal.
Date : 2026-01-10 Size : 5.35mb User : Alex
[DataMining] F_Score DL : 0: 特征选择的一种方法，对数据矩阵进行f打分，然后降序排序，可以从中发现重要的数据。-A method of feature selection, to f score data matrix, and then descending order, you can find important data.
Date : 2026-01-10 Size : 1kb User : Alex
[DataMining] Wavelet-Packet DL : 0: 基于混合信号的小波包分解技术在故障特征提取中的应用-Feature Extraction Using Multisignal Wavelet Packet Decomposition
Date : 2026-01-10 Size : 522kb User : 苏祥
[DataMining] 77257795PCA_yuandaima DL : 0: PCA源程序，主元分析源程序，可以用于变量的特征提取(PCA source code, principal component analysis source, can be used for variable feature extraction)
Date : 2026-01-10 Size : 2kb User : relief

« 12 »