随笔分类 -  机器学习

Machine Learning
CentOS 7安装tensorflow 1.6
摘要：1.6 依赖于cuda 9.0 cuDNN 7，并且前提需要下载和安装nvidia驱动。 安装驱动 nvidia 驱动地址： http://www.geforce.cn/drivers 添加： 然后执行： 重启，登录界面那里按ctrl+alt+f2，输入init 3。然后在命令行中执行那个run文件 阅读全文
posted @ 2018-03-18 01:26 linyx 阅读(1516) 评论(0) 推荐(0)
tensorflow bilstm官方示例
摘要：官方关于bilstm的例子写的很清楚了。因为是第一次看，还是要查许多东西。尤其是数据处理方面。 数据的处理(https://segmentfault.com/a/1190000008793389) 拼接 从shape的角度看： 抽取： 阅读全文
posted @ 2017-06-10 16:50 linyx 阅读(13364) 评论(1) 推荐(2)
ML | SVM
摘要：What's xxxAn SVM model is a representation of the examples as points in space, mapped so that the examples of the separate categories are divided by a... 阅读全文
posted @ 2014-07-20 15:54 linyx 阅读(252) 评论(0) 推荐(0)
ML| EM
摘要：What's xxxThe EM algorithm is used to find the maximum likelihood parameters of a statistical model in cases where the equations cannot be solved dire... 阅读全文
posted @ 2014-07-20 13:30 linyx 阅读(265) 评论(0) 推荐(0)
ML | Naive Bayes
摘要：what's xxxIn machine learning, naive Bayes classifiers are a family of simple probabilistic classifiers based on applying Bayes' theorem with strong (... 阅读全文
posted @ 2014-07-20 01:24 linyx 阅读(301) 评论(0) 推荐(0)
ML | k-means
摘要：what's xxxk-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, ... 阅读全文
posted @ 2014-07-19 23:24 linyx 阅读(170) 评论(0) 推荐(0)
ML | PCA
摘要：what's xxxPCAprincipal components analysis is for dimensionality reduction.主要是通过对协方差矩阵Covariance matrix进行特征分解，以得出数据的主成分（即特征向量eigenvector）与它们的权值（即特征值ei... 阅读全文
posted @ 2014-07-19 21:10 linyx 阅读(253) 评论(0) 推荐(0)
ML | spectral clustering
摘要：What's xxxIn multivariate statistics and the clustering of data, spectral clustering techniques make use of the spectrum (eigenvalues) of the similari... 阅读全文
posted @ 2014-07-19 18:26 linyx 阅读(247) 评论(0) 推荐(0)
Classical method of machine learning
摘要：PCAprincipal components analysiskmeansbayesspectral clusteringsvmEMhidden Markov modelsdeep learningSpark 阅读全文
posted @ 2014-07-19 14:49 linyx 阅读(157) 评论(0) 推荐(0)
Machine Learning Done Wrong【转】
摘要：1. Take default loss function for grantedMany practitioners train and pick the best model using the default loss function (e.g., squared error). In pr... 阅读全文
posted @ 2014-06-04 13:41 linyx 阅读(202) 评论(0) 推荐(0)