搜索资源列表
LapLDA
- 利用LapLDA进行初始化，预测出无标签数据XU的类别信息，这样所有的训练样本都有类别了。 为样本 的软类别标签，软类别标签中最大的问题对应其具体类别。 LapLDA可用基于原始数据的LLE-style reconstruction weights -LLE-style reconstruction weights
ML_Metric-function
- 上传的代码为多标签数据分类的度量函数 用于对多标签数据分类进行效果上的度量-Upload code for the multi-label data classification function is used to measure multi-label classification data to measure the effect on
mLknnMATLAB
- 本代码主要用了多标签K近邻方法（MLKNN）实现对多标签数据进行分类-This code mainly spent more than a label K-nearest neighbor method (MLKNN) to achieve the multi-label data classification
knnmyself
- knn分类算法，可以自动将excel转换为mat，可以自行调整参数，第二列为标签-knn classification algorithm can automatically be converted to excel mat, you can adjust the parameters of their own, as a second label
text-mining
- 文本挖掘，用词项-文档矩阵带入算法模拟出标签云和词条网络-text mining
IG
- 实现计算特征的信息增益，最后一列为样本标签，输出为特征的信息增益值及对应的序列号-Calculated to achieve gain characteristic information, and finally as a sample of tag output information characteristic of the gain value and the sequence number corresponding to
CHI
- 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
LDA
- LDA是监督式的降维算法，输入时需要为每一个数据打上标签信息。最多可以降到n-1维（n为数据点个数）-LDA Algorithm is used to realize dimensionality reduction. It can be used in the amount of projects such as face recognition.
DBNtoolbox-master
- 深度学习DBN（深信度网络）代码，概率生成模型，与传统的判别模型的神经网络相对，生成模型是建立一个观察数据和标签之间的联合分布。-Deep learning DBN (Convinced of the network) code generation probability model, and neural network models of traditional discrimination relatively generated model is to establish a join
pachongBDTB
- Python 爬去百度贴吧中一个贴子的内容，运用Urllib2和re模块，并对爬取的内容进行修改，去掉网页中的各种标签。-Python crawls the contents of a post in Baidu Post Bar, using Urllib2 and re modules, and crawl the contents of the amendment, remove the various pages of the label.
Sogou-character-porfile
- 介绍人物标签处理的过程，从数据采集，分词，预处理，算法选择以及结果展示方面来介绍相关过程。-This paper introduces the process of character label processing, and introduces the process of data acquisition, word segmentation, preprocessing, algorithm selection and result display.
lpa---java
- 标签传播算法(LPA)是由Zhu等人于2002年提出，它是一种基于图的半监督学习方法，其基本思路是用已标记节点的标签信息去预测未标记节点的标签信息。-Code label propagation algorithm
nlp7294
- 22w条打好标签的数据，供短文本主题分类使用(22W labeled data for short text topic classification)