Deep Learning(1) —— Andrew Ng

Binary Classification

把图像展平成一个列向量x,x作为输入得到输出y,y是一个判断是猫或不是猫的概率。

Notation used in this course

如果有m个训练样本,直观的做法可能是用for循环遍历所有的样本。但是在深度学习中应该像上图这样,把m个样本合成一个m列的向量(或矩阵),从而实现并行计算。

Logistic Regression

Sigmoid函数:\(\displaystyle \sigma(z) = \frac{1}{1+e^{-z}}\)

practice:

总结:

  • y是概率,通过y = wx + b这种线性回归的方法无法使y的值在0~1,于是我们引入Sigmoid函数,由线性回归变成逻辑回归。

Logistic Regression Cost Function

原文地址:https://www.cnblogs.com/pengweiblog/p/12625350.html

时间: 2024-10-28 01:52:35

Deep Learning(1) —— Andrew Ng的相关文章

Deep Learning by Andrew Ng --- PCA and whitening

这是UFLDL的编程练习.具体教程参照官网. PCA PCA will find the priciple direction and the secodary direction in 2-dimention examples. then x~(i)=x(i)rot,1=uT1x(i)∈R. is big when x(i)rot,2=uT2x(i) was small. so PCA drop x(i)rot,2=uT2x(i) approximate them with 0's. Whit

Deep Learning by Andrew Ng --- DNN

When should we use fine-tuning? It is typically used only if you have a large labeled training set; in this setting, fine-tuning can significantly improve the performance of your classifier. However, if you have a large unlabeled dataset (for unsuper

Deep Learning by Andrew Ng --- Softmax regression

这是UFLDL的编程练习. Weight decay(Softmax 回归有一个不寻常的特点:它有一个"冗余"的参数集)后的cost function和梯度函数: cost function: J(θ)=?1m??∑i=1m∑j=1k1{y(i)=j}logeθTjx(i)∑kl=1eθTlx(i)??+λ2∑i=1k∑j=0nθ2ij 梯度函数: ?θjJ(θ)=?1m∑i=1m[x(i)(1{y(i)=j}?p(y(i)=j|x(i);θ))]+λθj p(y(i)=j|x(i);

Deep learning by Andrew Ng --- Linear Decoder

Sparse Autoencoder Recap: Because we used a sigmoid activation function for f(z(3)), we needed to constrain or scale the inputs to be in the range [0,1], since the sigmoid function outputs numbers in the range [0,1].(sparse autoencoder输入层mean 0的原因).

(原创)Stanford Machine Learning (by Andrew NG) --- (week 7) Support Vector Machines

本栏目内容来源于Andrew NG老师讲解的SVM部分,包括SVM的优化目标.最大判定边界.核函数.SVM使用方法.多分类问题等,Machine learning课程地址为:https://www.coursera.org/course/ml 大家对于支持向量机(SVM)可能会比较熟悉,是个强大且流行的算法,有时能解决一些复杂的非线性问题.我之前用过它的工具包libsvm来做情感分析的研究,感觉效果还不错.NG在进行SVM的讲解时也同样建议我们使用此类的工具来运用SVM. (一)优化目标(Opt

(原创)Stanford Machine Learning (by Andrew NG) --- (week 10) Large Scale Machine Learning & Application Example

本栏目来源于Andrew NG老师讲解的Machine Learning课程,主要介绍大规模机器学习以及其应用.包括随机梯度下降法.维批量梯度下降法.梯度下降法的收敛.在线学习.map reduce以及应用实例:photo OCR.课程地址为:https://www.coursera.org/course/ml (一)大规模机器学习 从前面的课程我们知道,如果我们的系统是high variance的,那么增加样本数会改善我们的系统,假设现在我们有100万个训练样本,可想而知,如果使用梯度下降法,

Stanford CS229 Machine Learning by Andrew Ng

CS229 Machine Learning Stanford Course by Andrew Ng Course material, problem set Matlab code written by me, my notes about video course: https://github.com/Yao-Yao/CS229-Machine-Learning Contents: supervised learning Lecture 1 application field, pre-

神经网络作业: NN LEARNING Coursera Machine Learning(Andrew Ng) WEEK 5

在WEEK 5中,作业要求完成通过神经网络(NN)实现多分类的逻辑回归(MULTI-CLASS LOGISTIC REGRESSION)的监督学习(SUOERVISED LEARNING)来识别阿拉伯数字.作业主要目的是感受如何在NN中求代价函数(COST FUNCTION)和其假设函数中各个参量(THETA)的求导值(GRADIENT DERIVATIVE)(利用BACKPROPAGGATION). 难度不高,但问题是你要习惯使用MALAB的矩阵QAQ,作为一名蒟蒻,我已经狗带了.以下代核心部

Machine Learning Yearning - Andrew NG

链接(1~12章): https://gallery.mailchimp.com/dc3a7ef4d750c0abfc19202a3/files/Machine_Learning_Yearning_V0.5_01.pdf链接(第13章): https://gallery.mailchimp.com/dc3a7ef4d750c0abfc19202a3/files/Machine_Learning_Yearning_V0.5_02.pdf链接(第14章): https://gallery.mailc