Greetings all. I need good documentation about the development of decision trees and the mathematical algorithms used in building those trees. Also, what is the best program to use for this purpose? Is SPSS capable of running decision trees? If so, do I need to download any add-on to the basic SPSS program? Is it free to download? Finally, could this technique be used in engineering applications? Thanks in advance.
Decision trees are mainly used for classification problems. In this structure, each internal node corresponds to a predictive attribute (e.g., Age), each branch represents a test over its parent node (e.g., “is Age above 40?”) and each leaf consists of a class label. A new instance is classified through the application of successive tests that find a path in the tree from the root to some leaf node. In order to construct the tree, most algorithms adopt a greedy recursive strategy , where, at each step, some measure (e.g. entropy) is applied to define the attribute that best partitions the samples into distinct classes.
There are many toturials on the Internet. For instance, you may check: http://www.autonlab.org/tutorials/dtree18.pdf
I guess SPSS offers the CHAID algorithm for decision tree mining, but I don't think it is free to download.
There are many examples of decision trees applied to engineering problems. E.g.:
Perez, H. H., Datta-Gupta, A., Mishra, S., 2005. The Role of Electrofacies, Lithofacies, and Hydraulic Flow Units in Permeability Predictions from Well Logs: A Comparative Analysis Using Classification Trees. SPE Reservoir Evaluation & Engineering, 8 (2), SPE-84301-PA, 143–155.
See also the journal: "Engineering Applications of Artificial Intelligence"
- Badges
- Science topic
- Similar topics
- Mathematics