Also called Knowledge Discover in Databases (KDD), data mining is the process of automatically sifting through large volumes of data for patterns, using tools such as clustering, classification, association rule mining, and many more. There are several major data mining techniques developed and known today, and this article will briefly tackle them, along with tools for increased efficiency, including phone look up services.
Classification is a classic data mining technique. Based on machine learning, it is used to classify each item on a data set into one of predefined set of groups or classes. This method uses mathematical techniques, like linear programming, decision trees, neural network, and statistics. For instance, you can apply this technique in an application that predicts which current employees will most probably leave in the future, based on the past records of those who have resigned or left the company.
Association is one of the most used techniques, and it is where a pattern is discovered basing on a relationship of a specific item on other items within the same transaction. Market basket analysis, for example, uses association to figure out what products or services are purchased together by clients. Businesses use the data produced to devise their marketing campaign.
Sequential patterns, too, aim to discover similar patterns in data transaction over a given business phase or period. These findings are used for business analysis to see relationships among data.
Clustering makes useful cluster of objects that maintain similar characteristics using an automatic method. While classification assigns objects into predefined classes, clustering defines the classes and puts objects in them. Predication, on the other hand, is a technique that digs into the relationship between independent variables and between dependent and independent variables. It can be used to predict profits in the future - a fitted regression curve used for profit prediction can be drawn from historical sale and profit data.
Of course, it is highly important to have high-quality data in all these data mining techniques. A multi-database web service, for instance, can be incorporated to provide the most accurate telephone number lookup. It delivers real-time access to a range of public, private, and proprietary telephone data. This type of phone look up service is fast-becoming a defacto standard for cleaning data and it communicates directly with telco data sources as well.
Phone number look up web services - just like lead, name, and address validation services - help make sure that information is always fresh, up-to-date, and in the best shape for data mining techniques to be applied.
Source: http://ezinearticles.com/?Various-Data-Mining-Techniques&id=6985662
Classification is a classic data mining technique. Based on machine learning, it is used to classify each item on a data set into one of predefined set of groups or classes. This method uses mathematical techniques, like linear programming, decision trees, neural network, and statistics. For instance, you can apply this technique in an application that predicts which current employees will most probably leave in the future, based on the past records of those who have resigned or left the company.
Association is one of the most used techniques, and it is where a pattern is discovered basing on a relationship of a specific item on other items within the same transaction. Market basket analysis, for example, uses association to figure out what products or services are purchased together by clients. Businesses use the data produced to devise their marketing campaign.
Sequential patterns, too, aim to discover similar patterns in data transaction over a given business phase or period. These findings are used for business analysis to see relationships among data.
Clustering makes useful cluster of objects that maintain similar characteristics using an automatic method. While classification assigns objects into predefined classes, clustering defines the classes and puts objects in them. Predication, on the other hand, is a technique that digs into the relationship between independent variables and between dependent and independent variables. It can be used to predict profits in the future - a fitted regression curve used for profit prediction can be drawn from historical sale and profit data.
Of course, it is highly important to have high-quality data in all these data mining techniques. A multi-database web service, for instance, can be incorporated to provide the most accurate telephone number lookup. It delivers real-time access to a range of public, private, and proprietary telephone data. This type of phone look up service is fast-becoming a defacto standard for cleaning data and it communicates directly with telco data sources as well.
Phone number look up web services - just like lead, name, and address validation services - help make sure that information is always fresh, up-to-date, and in the best shape for data mining techniques to be applied.
Source: http://ezinearticles.com/?Various-Data-Mining-Techniques&id=6985662
No comments:
Post a Comment