Evolutionary design of decision trees for medical application

Abstract

Decision trees (DT) are a type of data classifiers. A typical classifier works in two phases. In the first, the learning phase, the classifier is built according to a preexisting data (training) set. Because decision trees are being induced from a known training set, and the labels on each example are known the first step can also be referred to as supervised learning. The second step is when the induced classifier is used for classification. Usually, prior to the first step several steps should be performed to improve the accuracy and efficiency of the classification: data cleaning, redundancy elimination, and data normalization. Classifiers are evaluated for accuracy, speed, robustness, scalability, and interpretability. DTs are widely used for exploratory knowledge discovery where comprehensible knowledge representation is preferred. The main attraction of DTs lies in the intuitive representation that is easy to understand and comprehend. Accuracy, however, is dependent on the learning data. One of the methods to improve the induction and other phases in the creation of a classifier is the use of evolutionary algorithms. They are used because the classic deterministic approach is not necessarily optimal with regard to the quality, accuracy, and complexity of the obtained classifier. In addition to the description of different evolutionary DT induction approaches, this paper also presents multiple examples of evolutionary DT applications in the medical domain. © 2012 Wiley Periodicals, Inc. This article is categorized under: Algorithmic Development > Hierarchies and Trees Technologies > Classification

Evolutionary design of decision trees

“…It is interesting to see this result compared with other well‐known classification algorithms. The UCI repository reports the classification results (classification accuracy) for 16 different algorithms using the original train/test split,11 the same as we did in our experiment. The best classification accuracy was obtained with two adapted Naïve‐Bayes approaches (FSS NB achieved 85.95% and NBTree achieved 85.90%), followed by our solution at rank three with 85.63%.…”

“…In this section, we present a case study where the whole process of inducing a DT using EAs is explained. For the case study, we will use Adult dataset from UCI Machine Learning Repository 11. The Adult dataset was extracted in 1994 from census data of the United States.…”

“…Next section focuses on evolutionary construction of DTs, after which techniques for performance improvements are presented. Finally, the last section represents a case study of evolutionary DT construction for the University of California, Irvine (UCI) Adult dataset 11…”

Evolving decision trees with beam search-based initialization and lexicographic multi-objective evaluation

“…A more recent strategy for avoiding greedy decision-tree induction is to generate decision trees through EAs [2,60,41]. We will focus the discussion of these methods regarding the initialization of individuals and the fitness evaluation strategies found in the literature, since our contributions are concerned mainly with these two aspects of EAs for decision-tree induction.…”

Revisiting evolutionary algorithms in feature selection and nonfuzzy/fuzzy rule based classification

“…Nowadays, the application domain of data mining is getting more complex and complex, i.e., it is shifted from traditional scientific26 and market basket database mining27 to biological,28,29 health care,30,31 agriculture,32 process monitoring and control,33 intrusion detection,34–36 and social network analysis 37. For example, detecting unauthorized use, misuse, and attacks that have no previously described patterns on information systems is usually a very complex task for traditional methods.…”