Data mining

It is difficult to re-estimate the meaning of the data which we collect continuously in the process of our work, to the business management or the manufacture management, to the banking capital, or to solve scientific, engineering and medical problems. The available data, in itself, is not enough for improving the work. We need to be able to transform the raw data into information which is useful for taking important business decisions. This is the main purpose of the Data mining technology. The phrase “knowledge discovery in databases” is often seen together with Data Mining. The goal of Data Mining is discovering hidden rules and lows in a dataset. The point is that the human mind in itself is not adapted to apprehend large quantity of heterogeneous information. In addition the human is not able to catch more than two or three relations even in not too big examples.

Data mining is the process of identifying valid, novel, potentially useful, and ultimately comprehensible knowledge from databases that is used to make crucial business decisions”

Gregory Piatetsky-Shapiro

Statistics & Reporting	Data Warehousing	OLAP	Data Mining	Business Intelligence
1970s	1980s	1990s	1990s

Data mining algorithms

While OLAP allows you to make analysis on present or past data, Data mining in fact allows making predictable analysis, based on present or past data. By Data mining you can ask and answer the following question: ”What will be our planned sales of hardware in Northern-East Region during the first quarter of the next year?”. The following table shows examples of task formulations by using OLAP and Data Mining methods:

OLAP	Data mining
What are the average indicators of traumatizing with smokers and non-smokers?	Which factors predict accidents best?
What are the average accounts of the existing customers comparing with the accounts of former customers (which don’t use already the services of the telephone company)?	Which characteristics distinguish the customers, which, most likely intend declining the services of the company?
What is the average of daily purchases by stolen or non-stolen credit card?	What schemes of purchases are typical of swindle with credit card?

The Data mining field is not limited. It can be everywhere, if there is any data. BI2M offers two Data mining algorithms - Clustering and Decision trees.