Home > Data Mining and Oracle Data... > What is Data Mining?
Data mining is the process of extracting useful information from masses of data by extracting patterns and trends from the data.
Data mining has it roots in machine learning, Machine learning, a discipline of artificial intelligence, deals with techniques that enable devices to learn from their own performance and modify their own functioning. Devices learn by building and modifying models. Data mining applies machine learning concepts to data.Machine learning identifies two kinds of learning:
Supervised learning, which builds predictive models
Unsupervised learning, which detects patterns
Data mining can be used to solve many kinds of business problems, including:
Predict individual behavior, for example, the customers likely to respond to a promotional offer or the customers likely to buy a specific product (Classification)
Find profiles of targeted people or items (Classification using Decision Trees)
Find natural segments or clusters (Clustering)
Identify factors more associated with a target attribute (Attribute Importance)
Find events or purchases that are likely to happen at the same time (Association, sometimes known as Market Basket Analysis)
Find fraudulent or rare events (Anomaly Detection)
Data mining is not restricted to solving business problems. For example, data mining can be used in the life sciences to discover the factors involved with a disease, to discover gene and protein targets, and to identify leads for new drugs.
Data mining builds models to detect patterns and trends. Building models is part of a larger process that starts with defining the problem to be solved and ends with deploying the model in a production environment. The data mining process is described in How Do You Mine Data?