Prediction for Traffic Accident Severity: Comparing the Artificial Neural Network, Genetic Algorithm, Combined Genetic Algorithm and Pattern Search Methods

Abstract

This paper focuses on predicting the severity of freeway traffic accidents by employing twelve accident-related parameters in a genetic algorithm (GA), pattern search and artificial neural network (ANN) modelling methods. The models were developed using the input parameters of driver's age and gender, the use of a seat belt, the type and safety of a vehicle, weather conditions, road surface, speed ratio, crash time, crash type, collision type and traffic flow. The models were constructed based on 1000 of crashes in total that occurred during 2007 on the Tehran–Ghom Freeway due to the fact that the remaining records were not suitable for this study. The GA evaluated eleven equations to obtain the best one. Then, GA and PS methods were combined using the best GA equation. The neural network used multi-layer perceptron (MLP) architecture that consisted of a multi-layer feed-forward network with hidden sigmoid and linear output neurons that could also fit multi-dimensional mapping problems arbitrarily well. The ANN was applied during training, testing and validation and had 12 inputs, 25 neurons in the hidden layers and 3 neurons in the output layer. The best-fit model was selected according to the R-value, root mean square errors (RMSE), mean absolute errors (MAE) and the sum of square error (SSE). The highest R-value was obtained for the ANN around 0.87, demonstrating that the ANN provided the best prediction. The combination of GA and PS methods allowed for various prediction rankings ranging from linear relationships to complex equations. The advantage of these models is improving themselves adding new data.

The Use of an Artificial Neural Network to Predict Australia’s Export Air Cargo Demand

“…The number of each layer of the ANN are shown in Figure 4 as a superscript on the variable of interest. This followed the approach of (Kunt et al, 2011) and (Srisaeng et al, 2015) where the authors used superscripts for identifying the source (second index) and also the destination (first index) for the various weights and other elements of the ANN network.…”

“…The elements of layer 1 (input layer), for instance, its bias, net input and output, have been assigned superscript 1. This indicates that they were associated with the input layer (layer 1) (Kunt et al, 2011).…”

“…Over-fitting can lead to predictions that are beyond the range of the training data (Richter and Weber, 2013). To avoid model over-fitting, the collected data was randomly divided into a 70:15:15 ratio (Garrido et al, 2014;Kunt et al, 2011;Srisaeng et al, 2015). A cross validation process was carried out during the training phase to avoid over-fitting of the proposed model (Efendigil et al, 2009).…”

“…The plot in Figure 6 has three lines, because all input and target vectors were randomly divided into three sets (Garrido et al, 2014;Kunt et al, 2011). As previously noted, 70 per cent of the vectors were used for training the network.…”

“…Hence, no formulation or a priori model is required (Merkus and Meesters, 2013;Watts and Worner, 2008). An ANN can be trained to perform a particular function by adjusting the values of connections (weights) between the model's elements (Kunt et al, 2011). During the training process, the ANN can detect complex relationships between the input and output data and perform synthesis (Sineglazov et al, 2013).…”