The power of decision tables

Abstract

Abstract. We evaluate the power of decision tables as a hypothesis space for supervised learning algorithms. Decision tables are one of the simplest hypothesis spaces possible, and usually they are easy to understand. Experimental results show that on artificial and real-world domains containing only discrete features, IDTM, an algorithm inducing decision tables, can sometimes outperform state-of-the-art algorithms such as C4.5. Surprisingly, performance is quite good on some datasets with continuous features, indicating that many datasets used in machine learning either do not require these features, or that these features have few values. We also describe an incremental method for performing crossvalidation that is applicable to incremental learning algorithms including IDTM. Using incremental cross-validation, it is possible to cross-validate a given dataset and IDTM in time that is linear in the number of instances, the number of features, and the number of label values. The time for incremental cross-validation is independent of the number of folds chosen, hence leave-one-out cross-validation and ten-fold cross-validation take the same time.

Relevance as a Metric for Evaluating Machine Learning Algorithms

“…We further investigate the influence of α and β parameters of the RS metric function. The performance of RS metric is studied through various experimentations using the Weka [10] simulator on the following rule-based prediction models: DecisionTable [11], JRip [12], Nearest Neighbor with generalization (NNge) [13], PART [14], ConjunctiveRule [15] and Ridor [16] on the breakout dataset [1]. The breakout dataset consists of 236 samples of data from different users gathered Relevance As a Metric for Evaluating Machine Learning Algorithms 9 from the breakout area.…”

Computational Estimation of Football Player Wages

“…The supervised machine learning methods used in the experiments cover several different machine learning paradigms, and include Additive Regression (Friedman, 2002), Decision Table (Kohavi, 1995), Nearest Neighbor with a weighted condition (Aha et al, 1991), K* (Cleary et al, 1995), Locally Weighted Learning with Naive Bayes and Linear regression classifiers (Frank et al, 2002;Atkeson et al, 1997), Random Committee (Seung et al, 1992), and Random Trees (Aldous, 1993).…”

“…Decision Table (Kohavi, 1995) is a rule-based learning algorithm that depends on testing a set of data using conditional mathematics, and predict the values of new samples using the results of the training set. Nearest Neighbor with a weighted condition (Aha et al, 1991) is an instance-based method that predicts the outcome of a given situation using the distance equation to find the nearest dataset to predict the results.…”

Performance of Classification Techniques on Medical Datasets

“…DecisionTable: Decision Table is an accurate method for numeric prediction from decision trees and it is an ordered set of If-Then rules that have the potential to be more compact and therefore more understandable than the decision trees [39].…”