Handling Imbalanced Data in Road Crash Severity Prediction by Machine Learning Algorithms

Abstract

Crash severity is undoubtedly a fundamental aspect of a crash event. Although machine learning algorithms for predicting crash severity have recently gained interest by the academic community, there is a significant trend towards neglecting the fact that crash datasets are acutely imbalanced. Overlooking this fact generally leads to weak classifiers for predicting the minority class (crashes with higher severity). In this paper, in order to handle imbalanced accident datasets and provide a better prediction for the minority class, the random undersampling the majority class (RUMC) technique is used. By employing an imbalanced and a RUMC-based balanced training set, we propose the calibration, validation, and evaluation of four different crash severity predictive models, including random tree, k-nearest neighbor, logistic regression, and random forest. Accuracy, true positive rate (recall), false positive rate, true negative rate, precision, F1-score, and the confusion matrix have been calculated to assess the performance. Outcomes show that RUMC-based models provide an enhancement in the reliability of the classifiers for detecting fatal crashes and those causing injury. Indeed, in imbalanced models, the true positive rate for predicting fatal crashes and those causing injury spans from 0% (logistic regression) to 18.3% (k-nearest neighbor), while for the RUMC-based models, it spans from 52.5% (RUMC-based logistic regression) to 57.2% (RUMC-based k-nearest neighbor). Organizations and decision-makers could make use of RUMC and machine learning algorithms in predicting the severity of a crash occurrence, managing the present, and planning the future of their works.

Developing Crash Severity Model Handling Class Imbalance and Implementing Ordered Nature: Focusing on Elderly Drivers

“…It focuses on how to predict minority classes more accurately by controlling the false positive rate increased. One of the solutions is based on the sampling strategy, which is broadly categorized by undersampling and oversampling [ 16 ]. The former (i.e., undersampling) is a sampling approach that reduces the size of a majority class so as to be “balanced” with that of a minority class, whereas the latter (i.e., oversampling) is to duplicate a minority class to increase its size.…”

“…Thus, the classifiers tend to predict majority classes more accurately than minority counterparts [ 14 , 15 ]. As Fiorentini and Losa (2020) [ 16 ] pointed out, most research works predicting crash severity overlooked the imbalanced class problem, leading them to develop and compare crash severity prediction models with or without handling the imbalanced problem. The authors recommended addressing the imbalanced issue when predicting crash severity.…”

Traffic Crash Severity Prediction—A Synergy by Hybrid Principal Component Analysis and Machine Learning Models

“…Although the accuracy was found to be very low, the authors did not apply any dimension reduction technique such as principal component analysis (PCA) on the crash dataset to overcome the problems of correlation between the input variables. Fiorentini and Losa [ 24 ] investigated the effect of applying balancing techniques on crash data on the performance of multiple ML models such as random tree, KNN, LR, and RF. It was found that introducing balancing technique enhanced the prediction power of the developed models.…”

Exploring the Injury Severity Risk Factors in Fatal Crashes with Neural Network

“…ML methods are more flexible with no or fewer model assumptions for input variables, and also have better fitting characteristics. Some of the commonly used ML approaches used in crash injury severity prediction include artificial neural networks (ANN) [ 58 , 59 , 60 ], random forest [ 54 , 61 , 62 ], support vector machines (SVM) [ 51 , 63 , 64 ], naïve Bayes [ 65 , 66 , 67 ], K-means clustering (KC) [ 68 , 69 , 70 ], and decision trees (DT) [ 71 , 72 , 73 ].…”