Data Mining（9.5第一堂课）

综述

成绩组成

1. 10%签到
2. 50%作业（在线动手作业，5次课）
3. 40%考察，期末作业。（阿里天池大数据竞赛）（https://tianchi.aliyun.com/competition/gameList/activeList）
   提交：英文报告（0.1）、代码（0.3）、数据集
4. 0-5分presentation
   

五个专题

Understanding your data

description and preparation（数据预处理）

Typical data mining tasks

1. pattern mining（模式挖掘）
2. Classification（分类）
3. Cluster analysis（聚类分析）
4. Outlier detection（异常值检测）

Introduction

Why

1. 针对大数据，数据的产生、存储、处理，越来越方式多样。
2. 数据挖掘，从数据到知识。价值来源于数据。

Evolution

1. before 1600，empirical
2. 1600-1950s，theoretical science
3. 1950s-1990s，computational science
4. 1900-now，data science

What Is Data Mining?

1. Data mining (knowledge discovery from data)
   Extraction of interesting (non-trivial, implicit, previously unknown and potentially useful) patterns or knowledge from huge amount of data
2. Alternative names: Knowledge discovery ( mining) in databases( KDD)

KDD Process

Data Cleaning>>>Data Warehouse >>>Task-relevant >>> Data Mining >>> Patern Evaluation >>> Knowledge

A Web Mining Framework

1. Web mining usually involves
2. Data cleaning
3. Data integration from multiple sources
4. Warehousing the data
5. Data cube construction
6. Data selection for data mining
7. Data mining
8. Presentation of the mining results
9. Patterns and knowledge to be used or stored into
10. knowledge-base