The easy-to-integrate NAG Data Mining Components are statistical and artificial intelligence software components specifically developed for the rigors of mining massive datasets. Robust, reliable and accurate algorithms are included in the Data Mining Components for each stage of the modeling process -data preparation (case wise deletion and dummy variables generation), data transformation (principle component analysis and data scaling), and model building (decision tree analysis; K-means and hierarchical clustering; multi-layer perceptron neural networks; logistic regression; general multiple regression). Application developers can use these components in their own application designs using standard development tools.
The NAG Data Mining Components are geared to help developers select the most appropriate modeling methods quickly and easily. Some of the algorithms can be called in a "novice" mode to get results faster or in "expert" mode to provide precise control of the algorithm. Less experienced users can also get the benefit of detailed advice via hyperlinked documentation guiding them to data mining solutions most applicable to their data.
The NAG Data Mining Components are designed to be incorporated into the user's application rather than requiring the user to learn a new interface to gain access to additional techniques. Closely integrating NAG Data Mining Components allows application developers to get exactly what the users need rather than accept the available results provided by other commercially available data mining tools.