J48PartiallyConsolidated: Class for generating a Partially Consolidated Tree-Bagging (PCTBagging) multiple classifier.
| URL: | | http://www.aldapa.eus/res/weka-pctbagging |
| Author: | | Jesús M. Pérez <txus.perez{[at]}ehu.eus>, Josué Cabezas Regoyo, Ander Otsoa de Alda Alzaga <ander.otsoadealda{[at]}gmail.com> |
| Maintainer: | | Jesús M. Pérez <txus.perez{[at]}ehu.eus> |
Class for generating a Partially Consolidated Tree-Bagging (PCTBagging) multiple classifier.
Allows building a classifier between a single consolidated tree (100%), based on J48Consolidated, and a bagging (0%), according to the given consolidation percent.
First, a partially consolidated tree is built based on a set of samples, and then, a standard J48 decision tree is developed from the leaves of the tree related to each sample, as Bagging does, combining J48Consolidated's interpretability with Bagging's discriminative power.
Key improvements:
- Flexible node specification: exact count or percentage of nodes
- Iterative construction that stops when reaching target nodes
- 7 node-selection criteria (Size, GainRatio variants, etc.)
- 2 search heuristics (Best-first and Hill climbing)
- Optional 'unrestricted' pruning of base trees
- Detailed node annotations showing development order and ensemble retention
- 40+ performance metrics for comprehensive evaluation
For more information, see:
Jesús M. Pérez and Olatz Arbelaitz.
"Multi-Criteria Node Selection in Direct PCTBagging: Balancing Interpretability and Accuracy with Bootstrap Sampling and Unrestricted Pruning". Information Sciences (2025), submitted.
doi:10.1016/j.ins.2025.XX.XXXAlso see:
Igor Ibarguren and Jesús M. Pérez and Javier Muguerza and Olatz Arbelaitz and Ainhoa Yera.
"PCTBagging: From Inner Ensembles to Ensembles. A trade-off between Discriminating Capacity and Interpretability". Information Sciences (2022), Vol. 583, pp 219-238.
doi:10.1016/j.ins.2021.11.010
All available versions:
Latest
1.1
0.3