The magnitude of data about the organization normally exceed the individual capabilities of its analysis, this is why it is needed to use the automated processes of data exploration .
Data Mining is a process of extracting knowledge from databases or data warehouses. The advanced algorithms in the field of artificial intelligence, such as statistical methods, neural networks, evolutionary methods and fuzzy logic are to analyze the data and to reveal new, unseen dependencies, patterns and knowledge. Such predictive treatment makes it possible to predict the behavior of organizations and market, to shape a business, to forecast the future, and to make plans.
POSSIBLE APPLICATIONS:
- segmentation, classification and better understanding of customers
- sales forecasting
- fraud detection
- detection of data dependencies
EXAMPLE: a data analysis in a grocery discount allows for detecting the following correlations:
- young men under 35 years old who buy chips on Thursday and Friday, usually buy also at least 2 cans of beer
- these men buy only mid-range beer
- on Friday they make large purchases, but on Thursday they buy only few products
- during summertime they buy also fruits, but only if they used to buy a lot of fruit juice and fruit preserves in the past
