DTREG is a robust software for statistical analysis that creates decision trees to model and forecast data. It generates Single-tree, TreeBoost, and Decision tree Models for classification and regression purposes.
DTREG is a versatile program that can build both Classification Trees and Regression Trees. With the former, the target variable being predicted is categorical, and the latter, the target variable is continuous, such as income or sales volume. The program has a simple interface where the user can easily check a button to direct DTREG to build a classic single-tree model, a TreeBoost model consisting of a series of trees or a Decision Tree Forest model.
DTREG is unique in that it uses V-fold cross-validation to determine the optimal tree size, thus eliminating the risk of overfitting. Overfitting occurs when the generated tree fits the training data well but does not provide accurate predictions for new data. Therefore, DTREG ensures that the tree size is optimal for accurate predictions.
Furthermore, DTREG uses a sophisticated technique involving "surrogate splitters" to handle missing values, and it enables DTREG to predict the values of cases with missing data. The technique ensures that cases with available data and some missing data are used to the maximum extent when building the model, making DTREG more reliable and comprehensive.
DTREG is unparalleled when it comes to displaying the generated decision tree. Users have the option to view the tree on the screen, have it written to a .jpg or .png disk file, or even print it. When printed, DTREG uses its advanced technique to paginate trees that cross multiple pages, making them easy to read and study.
Overall, DTREG is an outstanding software that is easy to use and has advanced features. It is a reliable tool that can create precise and accurate decision trees that can make complex data analyses and predictions in minutes.
Version 3.0: N/A