Dancing with Dirty Road Traffic Accidents Data: The Case of Gauteng Province in South Africa

Abstract

This article proposes an effective tree-based estimation method for dealing with the incomplete road traffic accidents (RTA) data problem. The essence of the approach is the proposal that RTA prediction can be improved by “doing more” with the missing data rather than discarding it. The author evaluates the approach with seven modern classifiers that have missing value strategies including C4.5, classification and regression trees (CART), k-nearest neighbour (k-NN), linear discriminant analysis (LDA), naïve Bayes classifier (NBC), repeated incremental pruning to produce error reduction (RIPPER) and support vector machines (SVMs). Experimental results are provided to illustrate the efficiency and the robustness of the proposed algorithm.