TY - EJOU AU - Malik, Meenakshi AU - Nandal, Rainu AU - Dalal, Surjeet AU - Jalglan, Vivek AU - Le, Dac-Nhuong TI - Driving Pattern Profiling and Classification Using Deep Learning T2 - Intelligent Automation \& Soft Computing PY - 2021 VL - 28 IS - 3 SN - 2326-005X AB - The last several decades have witnessed an exponential growth in the means of transport globally, shrinking geographical distances and connecting the world. The automotive industry has grown by leaps and bounds, with millions of new vehicles being sold annually, be it for personal commuting or for public or commodity transport. However, millions of motor vehicles on the roads also mean an equal number of drivers with varying levels of skill and adherence to safety regulations. Very little has been done in the way of exploring and profiling driving patterns and vehicular usage using real world data. This paper focuses on extracting and classifying distinct driving patterns using actual, dynamic vehicular data collected from the “On Board Diagnostics” port present (by default) in most vehicles. “Machine learning” and “Deep Learning” techniques were thereafter employed to extract and derive insights from observed patterns in the data. Various algorithms like hierarchical clustering, k-means clustering, multinomial naive bays, artificial neural networks and multi-layer perceptron were used to construct models to extract driving patterns and classify the data and ultimately generate insights about driver behavior across various parameters. The Inter-Class-ReLU was used to generate activation functions for the production of logical neurons and to develop and present a model that can classify incoming data into various groups and precisely identify distinct driving patterns. The developed model can be utilized in public as well as personal vehicles for monitoring driving behavior and driver preferences so that irregular behavior of a certain driver of a particular vehicle for a specified period can be scrutinized with ease. KW - On board diagnostics; hierarchical clustering; k-means clustering; multinomial Naive Bayes; artificial neural networks; multilayer perceptron DO - 10.32604/iasc.2021.016272