On the automatic design of decision-tree induction algorithms

  title={On the automatic design of decision-tree induction algorithms},
  author={Rodrigo C. Barros},
DEcision-tree induction is one of the most employed methods to extract knowledge from data. There are several distinct strategies for inducing decision trees from data, each one presenting advantages and disadvantages according to its corresponding inductive bias. These strategies have been continuously improved by researchers over the last 40 years. This thesis, following recent breakthroughs in the automatic design of machine learning algorithms, proposes to automatically generate decision… 
1 Citations

Automated design of genetic programming of classification algorithms.

The hypothesis that automating the design of GP classification algorithms for data classification can still lead to the induction of effective classifiers is investigated, and the automated designed classifiers were found to outperform the manually designed GP classifiers on all the problems considered in this study.



Towards the automatic design of decision tree induction algorithms

This work proposes two different approaches for automatically generating generic decision tree induction algorithms based on the evolutionary algorithms paradigm, which improves solutions based on metaphors of biological processes.

A hyper-heuristic evolutionary algorithm for automatically designing decision-tree algorithms

This work proposes a hyper-heuristic evolutionary algorithm for automatically generating decision-tree induction algorithms, named HEAD-DT, and shows that it can generate algorithms that significantly outperform C4.5 and CART regarding predictive accuracy and F-Measure.

Evolutionary Design of Decision-Tree Algorithms Tailored to Microarray Gene Expression Data Sets

A hyper-heuristic evolutionary algorithm called HEAD-DT is proposed that evolves design components of top-down decision-tree induction algorithms and is capable of generating algorithms that significantly outperform the baseline manually designed decision-Tree algorithms regarding predictive accuracy and F-measure.

A beam search based decision tree induction algorithm

The authors of this chapter present a new algorithm that seeks to avoid being trapped in local-optima by doing a beam search during the decision tree growth by keeping the comprehensibility of the traditional methods and is much less time-consuming than evolutionary algorithms.

Automatic Design of Decision-Tree Algorithms with Evolutionary Algorithms

This study reports the empirical analysis of a hyper-heuristic evolutionary algorithm that is capable of automatically designing top-down decision-tree induction algorithms, HEAD-DT, which is extensively tested using 20 public UCI datasets and 10 microarray gene expression datasets.

On the induction of decision trees for multiple concept learning

This dissertation makes four contributions to the theory and practice of the top-down non-backtracking induction of decision trees for multiple concept learning, and analyzes the merits and limitations of using the entropy measure (and others from the family of impurity measures) for attribute selection.

A Survey of Evolutionary Algorithms for Decision-Tree Induction

This paper presents a survey of evolutionary algorithms that are designed for decision-tree induction, which provides an up-to-date overview that is fully focused on evolutionary algorithms and decision trees and does not concentrate on any specific evolutionary approach.

Lexicographic multi-objective evolutionary induction of decision trees

A new GA-based algorithm based on a lexicographic multi-objective approach for decision tree induction that is able to avoid the previously described problems, reporting accuracy gains and induced models with a significantly reduction in the complexity considering tree sizes.

Evolution of Decision Trees

This work proposes to induce a decision trees (without regarding the type) with an unified algorithm based on artificial evolution, suggesting that Evolutionary Algorithms are competitive and robust for inducing all kinds of decision trees, achieving sometimes better performance than traditional algorithms.