lFIT: an unsupervised discretization method based on the Ramer–Douglas–Peucker algorithm

Abstract

Discretization is the process of converting continuous values into discrete values. It is a preprocessingstep of several machine learning and data mining algorithms and the quality of discretization may drastically affectthe performance of these algorithms. In this study we propose a discretization algorithm, namely line fitting-baseddiscretization (lFIT), based on the Ramer–Douglas–Peucker algorithm. It is a static, univariate, unsupervised, splittingbased, global, and incremental discretization method where intervals are determined based on the Ramer–Douglas–Peucker algorithm and the quality of partitioning is assessed based on the standard error of the estimate. To evaluatethe performance of the proposed method, a set of experiments are conducted on ten benchmark datasets and the achievedresults are compared to those obtained by eight state-of-the-art discretization methods. Experimental results show thatlFIT achieves higher predictive accuracy and produces less number of inconsistency while it generates larger numberof intervals. The obtained results are also validated through Friedman’s test and Holm’s post hoc test which revealedthe fact that lFIT produces discretization schemes that statistically comply both with supervised and unsuperviseddiscretization methods.