# Linear Classification

### Loss Function

Now we want to solve a image classification problem, for example classifying an image to be cow or cat. The machine learning algorithm will score a unclassified image according to different classes, and decide which class does this image belong to based on the score. One of the keys of the classification algorithm is designing this loss function.

Map/compute image pixels to the confidence score of each class

Assume a training set:

$(x_i,y_i)$

$$x_i$$ is the image and $$y_i$$ is the corresponding class

i∈1…N means the traning set constains N images

$$y_i$$∈1…K means there are K image categories

So a score function maps x to y:

$f(x_i,W,b)=W\cdot x_i+b$

In the above function, each image $$x_i$$ is flattend to a 1 dimention vector

If one image’s size is 32x32 pixels with 3 channels

$$x_i$$ will be a 1 dimention vector with the length of D=32x32x3=3072 Parameter matrix W has the size of [KxD], it is often called weights b of size [Kx1] is often called bias vector In this way, W is evaluating $$x_i$$’s confidence score for K categories at the same time

# 用Jekyll模板搭建Github页面-Windows

### 安装基本软件

• 首先安装一个能在windows环境下运行的包管理器Chocolatey

• 因为Jekyll是用Ruby写的，所以要安装Ruby，在控制台中输入choco install ruby -y回车

• 关闭控制台，然后再打开控制台并输入gem install jekyll，这样Jekyll就装好了：如果出现ssl3错误按照以下步骤（点我看原文）解决：

cmd输入 gem install –local C:\rubygems-update-x.x.xx.gem：local后面即刚下载好的gem文件

然后输入update_rubygems –no-ri –no-rdoc

结束后再输入gem install jekyll，应该就可以了

• 重新打开控制台，输入chcp 65001避免编码问题

• 安装Ruby开发环境，在控制台中输入：

choco install ruby2.devkit

• C:\tools\DevKit2文件夹中打开控制台，执行命令 ruby dk.rb init，产生config.yml文件

# 改造Jekyll模板的技术细节

#### 框架的文件夹结构

_layout ：主要定义了两种类型页面的排版，post是为单篇文章设计的排版，post-index是为一系列文章设计的排版。

_posts：用于存放所有文章的md文件，md文件的命名必须严格按照”年-月-日-标题”的格式命名。

_sass：用于存放定制的css文件，比如_page就规定了页面各个元素的宽度颜色字体，_variables定义了一些全局变量的值。

_site：模板编译完成后生成的页面，这个是真正可以直接部署的页面，平时不用看

_templates：规定了不同类型的排版文件中可以定义的变量

images：用于存放图片

search：用于存放搜索框页面

tags：用于存放按照tags列出所有文章的页面

categories：用于存放按照category列出所有文章的页面

posts：用于存放列出所有文章的页面

# Goodbye Mobis

After two years of happy and constructive time in camera team, I got some homesick, I decided to go back to my 2nd hometown, Munich.

The days in Mobis are amazing. I learned so many new things and read so many papers that I can not even believe it, especially when I am cleaning them for the moving, Luka’s enthusiasm of new technologies motivated me to track the trend, which at end, benefits the design of our technical roadmap. The lessons I learned from Luka worth dual, maybe triple master degrees, one for geometry, one for deep learning and one for management, together with the lessons learned from everyone of camera team, I will definitely give myself an honored PhD title, awarded by camera team.

Before I joined camera team, I only have friends from few countries, now I have worked with the talents from so many countries and cultural backgrounds, the experience is unbelievably great. I got to understand so many cultures and religions, vividly, in person. My view of the world expanded so much here that I understand the world much more than before. The different cultures didn’t shock we, the diversity and the harmony of our team do amazed me. I would give special thanks to Thusita for constructing such a great team.

Conflicts and different technical views do exist, which is always helpful for us to regularize each other, I would consider these more as a kind of unsupervised learning. Thank you everyone for your tolerance on me, there are so many s**ty words came from my mouth and I am quite sure some of you guys definitely got ear pollution or even ear cancel.

I feel sad that I am leaving the camera team at this specific time, by doing good work, we win the trust from MTCK, we are working on more and more projects, we are moving to a bigger place and expanding further, last but not least, we just got a new lunch supplier!

It has been a great pleasure working with you all and I believe that we will have successful collaboration in the future!