ML
文章平均质量分 85
Tongust
这个作者很懒,什么都没留下…
展开
-
ESL-ICA
Independent Component AnalysisICA Maximum LikelihoodSigmoid functionFastICAWhitenPython ImplementIndependent Component Analysis假设x∈ℝp×1x\in \mathbb{R}^{p\times 1}是观测直,s∈ℝp×1s\in \mathbb{R}^{p\tim原创 2017-04-27 01:20:04 · 691 阅读 · 0 评论 -
统计学习之---多尺度变换 MDS (Multidimensional Scaling)
Multidimensional Scaling (MDS)The idea is to find a set of k-dimensional continuous variables to representation the original data, which only contains the dis-similarity information. It is used in prob原创 2017-03-26 23:15:11 · 4451 阅读 · 0 评论 -
The solution on the Elements of Statistical Learning
统计学习基础 第二版 参考答案原创 2016-12-17 15:22:50 · 4251 阅读 · 0 评论 -
Iterative Reweighted Least Squares
逼近 approximation加权最小二乘weighted least squared error approximation逼近其他范式IRLS逼近 lpl_p normOverdetermined system N pAlgebra 方法IRLS for Logit RegressionModeling本文基于这篇paper。逼近 (approximation)对于线性的问题,原创 2017-05-03 21:27:21 · 2477 阅读 · 1 评论 -
Newto-Raphson
Newton-Raphson一阶形式推导过程二阶形式推导过程Convex FunctionQuadratic Program1. Newton-Raphson解决f(x)=0f(x)=0问题的逼近方式一阶形式xnew=xold−f(x)f′(x)f′(x)=∂f(x)∂xx^{new} = x^{old}-\frac{f(x)}{f'(x)}\\f'(x) = \frac{\par原创 2017-05-03 22:38:10 · 287 阅读 · 0 评论 -
Gap Statistic 间隔统计量
Gap Statistic聚类的紧支测度 measure of the compactness间隔统计量GSPython 实现测试Gap StatisticGap statistic由Tibshirani等人提出,用以解决聚类问题确定所判定类的数目。聚类的紧支测度 (measure of the compactness)最简单的方法是使用类内样本点之间的欧式距离来表示,记为DkD_k,DKD原创 2017-04-26 01:44:03 · 24195 阅读 · 7 评论 -
Kalman Filter
Kalman Filter1 Linear Optimal Filtering2 Orthogonality Principle4 State Model5 Objective And HypothesisObjectiveHypothesisRemark3 NotationsRemarkPropositionConclusion of Space RelationshipT原创 2017-05-21 15:33:40 · 456 阅读 · 0 评论