Data Preprocessing in Data Mining

reference: Chapter 3 in book <<Data Mining Concepts and Techniques>>, 3rd Ed by Han, etc.

Data preprocessing techniques aims to improve the quality of pattern mined and/or the time required for mining:

• Data cleaning, which removes noise and correct inconsistencies
• Data integration, which merges multiple data sources into a data store, i.e. data warehouse
• Data reduction, which reduces data size by i.e. aggregating, redundancy reduction, clustering, etc
• Data transformation, which scales data into a smaller range (i.e. normalization to 0 to 1)

The above techniques can work together. Data characteristics, attribute features are useful for these techs.