CCM Blog

Mac上用iTerm使用rz

安装lrzsz brew install lrzsz 下载iterm2-zmodem git clone https://github.com/mmastrac/iterm2-zmodem.git iTerm里面设置Trigger，路径：Preferences-Profiles-Advanced-Trigger Edit ...

Posted by c cm on June 12, 2017

Generative Adversarial Networks

生成模型：用从分布$p_{data}$采样得到的训练集，学习估计出该分布的模型。形式可以是估计出分布$p_{model}$，也可以是学会生成$p_{model}$的样本。 1 生成模型优点生成模型和强化学习关系密切。model-based强化学习中就包含生成模型。比如可用来学习根据现在state和action，生成未来state的分布；使用生成学习来模拟，还能减少因为错误acti...

Posted by c cm on March 15, 2017

Factorization Machines

1. 模型定义其中，k为描述分解维度的超参。优点：特征稀疏时可用线性复杂度 O(kn) 2. 学习直接优化primal 3. 模型对比 SVM 线性核SVM一致比Polynomial核，权重间是相关的比其他Factorization方法如Matrix and Tensor Factori...

Posted by c cm on March 3, 2017

Ad Click Prediction

1. FTRL $\textbf{g}t$ 第t个样本向量 $g{t, i}$ 第t个样本向量的第i个变量 $\textbf{g}_{1:t} = \sum_1^t \textbf{g}_i$ OGD(Online Gradient Descent): 其中$\eta$为学习速率。缺点为稀疏性不足。 FTRL: 其中$\sigma$为学习率，$\sigma_{1:t} ...

Posted by c cm on February 22, 2017

Dynamic Topic Models

1. Dynamic Topic Model 1. Static Topic Model eg. LDA $\beta_{1:K}$ K topics, multinomial distribution of fixed volcabulary Generative Process: Choose topic proportions θ from a distributio...

Posted by c cm on February 15, 2017

Recent Advances in Convolutional Neural Networks

1. Introduction 基本结构： input layer first layer convolutional layer aims to learn feature representations of the inputs. The activation function introduces nonline...

Posted by c cm on February 9, 2017

Python immutables异常行为

from copy import copy >>> a = [1, 2, 3] >>> b = a[:] >>> a is b # 1 >>> copy(a) is a # 2 >>> c = (1, 2, 3) >>> d = c[:] >>> e = (1, ...

Posted by c cm on December 14, 2016

贝叶斯规划学习

最新一期的《环球科学》将贝叶斯规划学习（Bayesian program learning，简称BPL）评为2016年十大创新之一。找到其源头Human-level concept learning through probabilistic program induction科普一下。人类学习的两个特点还没有被现在state-of-art的机器学习获得：大多数算法还是data-...

Posted by c cm on December 12, 2016

逻辑回归

基本模型（二元）对条件概率建模： y=1的对数几率(log odds/logit)为 $log \frac{p}{1 - p}＝w \cdot x$ 参数估计方法一般fit方式为极大似然法maximum likelihood，对其负对数似然函数（cross-entropy error）求极小值，就能得到参数$w$的估计值。具体求解可用梯度下降法（一阶...

Posted by c cm on December 4, 2016

knn

KNN是memory-based learning，或instance-based learning最简单的例子。在K为1时，可以生成Voronoi Diagram。除了一般的定义方式，即选择离所需预测点最近的k个点的某种描述来预测，还可以从KDE角度来看。 TBC

Posted by c cm on November 29, 2016