Cape Town road traffic accident analysis: Utilising supervised learning techniques and discussing their effectiveness

Abstract

[EN] Road traffic accidents (RTA) are a major cause of death and injury around the world. The use of Supervised learning (SL) methods to understand the frequency and injury-severity of RTAs are of utmost importance in designing appropriate interventions. Data on RTAs that occurred in the city of Cape Town during 2015-2017 are used for this study. The data contain the injury-severity (no injury, slight, serious and fatal injury) of the RTAs as well as several accident-related variables. Additional locational and situational variables were added to the dataset. Four training datasets were analysed: the original imbalanced data, data with the minority class over-sampled, data with the majority class under-sampled and data with synthetically created observations. The performance of different SL methods were compared using accuracy, recall, precision and F1 score evaluation metrics and based on the average recall the ANN was selected as the best performing model on the validation data.