Analysis and prediction of telecommunication data: DATA-MINING

A classification algorithm must explore structured semi-structured or unstructured informational data to detect their similarity and assign them the same category. Among the main applications we find drilling analysis and data mining. These areas of expertise which emerged less than a decade ago are attracting increasing interest from computer scientists mathematicians and even humanities specialists. The loss of a company's customers commonly known as attrition is a real problem for companies operating in different sectors of activity especially in competitive situations. It is well known that this phenomenon has not spared the telecommunications sector. To help decision makers to take a decision in front of all these questions we propose the design of a mini data warehouse (Data marts) to make an easy exploration of the historical data of the customers in connection with the profiles of uses of the customers based on their profiles. In our research we used the DATA MINING method to analyze and predict the data.