Data Mining and Machine Learning Techniques

The application of data mining and machine learning techniques in the highway safety analysis has boomed, resulting from the new and emerging data sources, powerful algorithms, handy software applications, and comparable or superior performance in crash prediction. The broad selection of techniques ranging from exploratory data analysis such as association rules, clustering analysis, decision tree models, Bayesian networks to more sophisticated neural network models, and support vector machines presents great opportunities to consider a large multitude of factors and explore intricate relationships among them. Most techniques are readily implemented through commercial or free statistical software packages such as R.