Types of Data Science Questions

In approximate order of difficulty

• Descriptive (描寫的；記述的)
• Exploratory (勘探的；探究的)
• Inferential (推理的；推論的)
• Predictive (預言性的；預兆的)
• Causal (原因的；因果的)
• Mechanistic (機械論的)

 

Types		Goal
Descriptive	描寫的；記述的	Describe a set of data
Exploratory	勘探的；探究的	Find relationships you didn’t know about
Inferential	推理的；推論的	Use a relatively small sample of data to say something about a bigger population
Predictive	預言性的；預兆的	To use the data on some objects to predict values for another object
Causal	原因的；因果的	To find out what happens to one variable when you make another variable change
Mechanistic	機械論的	Understand the exact changes in variables that lead to changes in other variables for individual objects

 

Descriptive (描寫的；記述的)

http://www.census.gov/2010census/

http://books.google.com/ngrams

 

Exploratory (勘探的；探究的)

http://www.sdss.org/

 

Inferential (推理的；推論的)

http://journals.lww.com/epidem/Abstract/2013/01000/Effect_of_Air_Pollution_Control_on_Life_Expectancy.4.aspx

 

Predictive (預言性的；預兆的)

http://fivethirtyeight.blogs.nytimes.com/

 

Causal (原因的；因果的)

http://www.nejm.org/doi/full/10.1056/NEJMoa1205037?query=featured_home&

 

Mechanistic (機械論的)

http://www.fhwa.dot.gov/resourcecenter/teams/pavement/pave_3pdg.pdf

R Packages

 

 

 

 

Machine Learning – What is machine learning?

What is Learning?

 

What is Skill?

 

How to know it is a Tree?

 

The machine learning route:

 

Key essence of machine learning:

 

Fun Time:

R Introduction

Robert Gentleman + Ross

1. Free Software

2. Lots of Users, more than 2 million users

3. Lots of Applications, more than 5000 add-on (2013/10)

 

Data 競賽時選用的語言

 

Number of Google Scholar Hits For Each Software

 

Number of Posts

 

R Environment Installation:

1. R-Core

2. R IDE

2.1 RStudio Desktop / RStudio Server (http://www.rstudio.com/)

2.2 Architect (Eclipse-based IDE)

2.3 Revolution R (Optimization R-Core) for Academics

 

R all applications/packages

1. CRAN Task Views: 任務分組, 使用目的分組

http://cran.r-project.org/web/views/

安裝群組套件之前，必須先安裝並載入ctv套件

install.packages(‘ctv’)

library(ctv)

install.views(‘The one you want to install’)

2. R-forge: 鍛造廠, 較前瞻性的會放這

https://r-forge.r-project.org/

 

R Knock Out Word!

1. Dynamic Documents

2. Reproducible Research

 

R 的文書功能:

1. Sweave (R + LaTex)

2. Knitr (R + Markdown)

3. IPython Notebook

 

R Input / Output Example:

1. Read local file (single/multiple)

2. Read network file

3. Read web page (SpideR)

4. Website dynamic interface (Shiny)

 

Code School, Try R:

https://www.codeschool.com/courses/try-r

 

R 推薦連結:

1. idre UCLA

2. Taiwan User R Group (Facebook)

3. Taiwan User R Group (Youtube)

4. 中華R軟體學會

5. 統計之都 COS

6. R Bloggers

7. R Language Resources (RevolutionAnalytics.com)

8. insider-R (美觀的R線上文件)