神经网络技术在统计建模、数据的转换分类回归等各大领域都有很大的应用空间。但是由于计算本身的复杂性以及早期的计算能力不足，神经网络一直没有得到很大的发展。然而近几年，随着GPU计算的进步，涌现出一大批非常强大而实用的神经网络的训练框架如Caffe、Keras、CUDA convnet、Torch7等。在这片教程中，我们着重介绍另一款灵活又好用的框架：Chainer的基础使用方法。你可以通过Jupyter Notebook或是其他的python终端来跟进这篇教程。
在这篇教程中，我们将先通过编写一个简单的线性回归器来帮助你入门，然后我们再编写一个用于识别MNIST手写数字的标准的深度学习模型来让你熟悉编程逻辑。
This article propose an object detection system that relies on a multi-region deep convolutional neural network that also encodes sematic segmentation-aware features. The module aims at capturing a diverse set of discriminative appearance factors and exhibits localization sensitivity that is essential for accurate object localization. They exploit the above properties of their recognition module by intergrating it on an iterative localization mechanism that alternates between socring a box proposal and refgining its location with a deep CNN regression model. And consiquently, they detect objects with very high localization accuracy.
Cause the wealth of information available on the web keeps growing, we want to use web crawler to get them. But the traditional method has a fatal limit of its large infrastructure cost. To reduce it, we developed this method, unicrawl, which can show a performance improvement of 93.6% in terms of network bandwidth consumption, and a speedup factor of 1.75.
Mobile devices have revolutionized how people share photos with each other on social networks with a single click of a button.
The content of the photos people shared to the internet are rarely analyzed by the websites before the photos are made available to view. After the photos are posted on the social network to the public to view, it is close to impossible to permanently delete the uploaded photos.
Photo leakage, regretion after posting and malicious posting happens from time to time.
版本信息：
Ubuntu15.10
Apache2.4.12
php5.6.11（zabbix3.0要求php版本至少5.4以上）
Mysql5.6.31
zabbix3.0
前言：本教程包括了ubuntu上LAMP(Linux+Apache+Mysql+Php)环境的搭建以及zabbix安装。因为我们最终是要通过外部计算机访问我们的服务器的，所以我希望你可以先运行一下“ifconfig -a”语句来查看以下自己的IP地址，以方便之后测试服务器。文中将以“IPAddr”来代替你的IP地址，阅读时请注意。
这是博主虚拟机上的IP地址：