---restore content starts---
After reading the big talk data mining this book the first 36 pages, learned the knowledge.
Data Mining (Mining) and Knowledge Discovery (KDD) in the database are aliases to each other.
Examples of data mining: beer and diapers, flow plan user base, package user churn reason, bundle sales, part maintenance fee moderation.
The concept of data mining: The discovery of hidden valuable knowledge based on large, incomplete, noisy, fuzzy, random data. The incomplete meaning is that in collecting the corresponding data there is missing, the noise is obtained by the data deviate from the real value, such as external interference, measuring instrument failure, manual input or copying error, etc. Fuzziness refers to the uncertainty of the concept of the subject itself, the reference, such as height. Randomness refers to the uncertainty of whether an event occurs or not, such as an impromptu idea of buying beer.
The functions of data mining: correlation analysis, cluster analysis, classification analysis, prediction, regression analysis, discriminant analysis, time series analysis, deviation screening and so on.
The three pillars of data mining: Database, statistics, machine learning.
Data warehousing: A topic-oriented, integrated, time-varying, persistent collection of data that supports the decision-making processes of the management layer.
Online analytical Processing,olap: an analytical technique that combines data, merges and aggregates, and observes information from different perspectives.
---restore content ends---
First day of data mining