Email: tmajumder.work@gmail.com
This repository contains several decision tree algorithms compatible with Scikit-Learn's Bagging Classifier. For the complete experimental setup and results, please check my thesis. If you find this code useful, please cite my work.
Majumder, T. (2020). Ensembles of oblique decision trees [Master's Thesis, University of Texas, Dallas]. UTD Theses and Dissertations.
Decision Trees considered for this experiment:
* Standard Decision Tree with Bagging
* Oblique Classifier 1 with Bagging
* Weighted Oblique Decision Tree with Bagging
* Randomized CART with Bagging
* HouseHolder CART with Bagging
* Continuous Optimization of Oblique Splits with Bagging
* Deep Neural Decision Trees with Bagging
* Non-Linear Decision Trees with Bagging
* Random Forest Classifier
In this experiment we have to skip OC1, DNDT and, NDT classifiers due to its computational cost.
This experiment has been conducted on 12 Benchmark Data sets.
* Iris * Vehicle
* Wine * Fourclass
* Glass * Segment
* Heart * Satimage
* Breast-cancer * Pendigits
* Diabetes * Letter