http://www.microsoft.com/sql/technol...dm/addins.mspx
The SQL Server 2005 Data Mining Algorithms
There are a number of algorithms available in SQL Server 2005 (Table 1).
Table 1. The algorithms featured in SQL Server 2005 Data Mining
Model Description
Decision Trees The Decision Trees algorithm calculates the odds of an outcome based on values in a training set. For example, a person in the age group 20-30 that makes over $60,000/year and owns a home is more likely to need a lawn service than someone in the age group of 15-19 who doesn't own a home. Based on age, income, and home ownership, the Decision Trees algorithm can calculate the odds of that person needing a lawn service based on historical values.
Association Rules The Association Rules algorithm helps identify relationships between various elements. For example, it is used in cross-selling solutions because it notes relationships between items and can be used to predict what else someone buying a product will also be interested in purchasing. The Association Rules algorithm can handle incredibly large catalogs, having been tested on catalogs of over half a million items.
Naive Bayes The Naive Bayes algorithm is used to clearly show the differences in a particular variable for various data elements. For example, the Household Income variable differs for each customer in the database, and can be used as a predictor of future purchasing. This model excels at showing the differences between certain groups such as customers who churn and those who don't.
Sequence Clustering The Sequence Clustering algorithm is used to group or cluster data based on a sequence of previous events. For example, users of a Web application can often follow a variety of paths through the site. This algorithm can group customers based on their sequence of pages through the site to help analyze users and determine if some paths are more profitable than others. This algorithm can also be used to predict, such as predicting the next page a user may visit. Note that the predictive capability of the Sequence Clustering algorithm is something that many other data mining vendors cannot deliver.
Time Series The Time Series algorithm is used to analyze and forecast time-based data. Sales are the most commonly analyzed and predicted data using the Time Series algorithm. This algorithm looks for patterns across multiple data series so that businesses can determine how different elements affect the analyzed series.
Neural Nets Neural networks are the core of artificial intelligence. They seek to uncover relationships in data that other algorithms miss. While the Neural Nets algorithm tends to be slower than the other algorithms, it finds relationships that may be non-intuitive.
Text Mining The Text Mining algorithm appears in SQL Server Integration Services and analyzes unstructured text data. This allows companies to analyze unstructured data such as a "comments" section on a customer satisfaction survey.
Extensibility
While SQL Server 2005 includes a number of algorithms out of the box, the model used by SQL Server 2005 allows any vendor to add new models into the data mining engine. Those models become peers with the models that come with SQL Server 2005. Algorithms from third parties also benefit from all the other features: they are callable using DMX and are easy to integrate into any part of the Integrate, Analyze, and Report process.
hello barnix
Linear Mode
