2014年08月_始于长虹

转载生成模型与判别模型的关系

一、决策函数Y=f(X)或者条件概率分布P(Y|X) 监督学习的任务就是从数据中学习一个模型（也叫分类器），应用这一模型，对给定的输入X预测相应的输出Y。这个模型的一般形式为决策函数Y=f(X)或者条件概率分布P(Y|X)。决策函数Y=f(X)：你输入一个X，它就输出一个Y，这个Y与一个阈值比较，根据比较结果判定X属于哪个类别。例如两类（w1和w2）分类问题，如果

2014-08-29 11:22:38 988

原创 Descriptor Matching with Convolutional Neural Networks: a Comparison to SIFT

论文笔记之《Descriptor Matching with Convolutional Neural Networks: a Comparison to SIFT》。

2014-08-26 15:11:38 3367 3

原创 Unsupervised feature learning by augmenting single images

学习笔记之《Unsupervised feature learning by augmenting single images》

2014-08-25 16:36:00 1680

原创 Learning and Transferring Mid-Level Image Representations using Convolutional Neural Networks

学习笔记之《Learning and Transferring Mid-Level Image Representations using Convolutional Neural Networks》

2014-08-21 16:17:08 4009 2

原创 Rich feature hierarchies for accurate object detection and semantic segmentation

学习笔记之《Rich feature hierarchies for accurate object detection and semantic segmentation》

2014-08-20 14:43:37 8918 2

原创 Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition

学习笔记之《Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition》

2014-08-15 17:22:44 5586

原创 Multi-Scale Orderless Pooling of Deep Convolutional Activation Features

学习笔记之《Multi-Scale Orderless Pooling of Deep Convolutional Activation Features》

2014-08-14 17:23:35 5144 3

原创 CNN: Single-label to Multi-label

学习笔记之《CNN: Single-label to Multi-label》

2014-08-13 15:40:07 9967 6

原创 Some Improvements on Deep Convolutional Neural Network Based Image Classif ication

学习笔记之《Some Improvements on Deep Convolutional Neural Network Based Image Classif ication》

2014-08-12 15:21:14 1979 1

woyaopojie1990的专栏