General Info
Home
News
CEO
Mail us
Contact
Links
Product Info
Shih Data Miner
Apps & Services
Screenshots
Technical Info
Download
ChangeLog
Manual
License
Recent users
Income rules
Motors case
Publications
Classification methods
Screenshots Shih Data Miner
| General view of the tree editor which contains the view of the tree, the splitting center, the pruning sequence and a small view of the tree that allows the user to easily explore the tree. |  |
The colors give the distribution of the class within a node in the tree. It easily indicates the classification of the nodes, in this case the whiter the node the more similar it is to the class you want to predict. |  |
|
| Of course, it is possible to extract more detailed information of the tree. This can be done by analyzing the lift and gain charts which show information of every terminal node. |  |
The model information is set on the model window. There you can set the training and testing files, the algorithm, the misclassification cost and many other parameters. |  |
|
| You have full control over the way to split a node. You can select the variable and the value to split and even the splitting algorithm or either you can let the node grow automatically. |
|
The train and test data are scored by the tree. A given example is classified according to its score. The model will classify an example according the class that scores higher on him. |
|
|
| In the splitting center it is possible to select the variable to split, and the value of the cut-off. Also it is possible to see the goodness of split for every possible value the modeler would chose and the class distribution of the train and test examples for the given node (selected in blue). |
|
The misclassification matrix or the confusion matrix gives aggregate information about how the model predicts versus the real data (for training and for testing data). This gives a very understandable indicator of how the model performs. |
|
