Turkish Journal of Physiotherapy and Rehabilitation; 32(2) ISSN 2651-4451 | e-ISSN 2651-446X www.turkjphysiotherrehabil.org 1024 EFFECTIVE RADICAL DRIVER SUPPORT SYSTEM USING MACHINE LEARNING METHODS FOR CONNECTED VEHICLES DR. K. S. BHUVANESHWARI 1 , DR. S. KANNIMUTHU 3 , 1,3 Associate Professor, Department of CSE, Karpagam College of Engineering, Coimbatore, Tamilnadu, India bhuvana.me@gmail.com, kannimuthu.s@kce.ac.in DR.D.BHANU 2 2 Professor, Department ofIT, Karpagam Institute of Technology bhanu.saran@gmail.com M. KARTHI 4 , K.V. HARSHA SAGAR 4 4 Research Student, Department of Mechanical Engineering, Karpagam College of Engineering, Coimbatore, Tamilnadu, India ABSTRACT: In recent days, connected vehicle technology is most dominating field in automotive industries and it is the most greatly researched area. The benefits of connected vehicles are realized in many aspects like crash elimination, improved energy efficiency, improved safety etc. Machine learning (ML) a subset of Artificial Intelligence that yields systems learning capability and progress from experience without the intervention of humans and without the explicit programming model. Applying machine learning algorithms in connected vehicles technology is need of an hour these days to effectively perform tasks in critical situations without human interventions. In this article, a machine learning framework model is proposed for the head pose estimation in connected vehicles. A comprehensive analysis on various machine learning algorithms is performed by using Advanced Driver Assistance System (ADAS) database. Experimental results show that decision tree algorithm outperforms the other learning algorithms well when compared to other supervised learning algorithms. Keywords:- Automotive industry, Connected Vehicles, Driving Performance, Machine Learning, Supervised learning I. INTRODUCTION Connected vehicle technology is likely to play a vital role in minimizing the occurrences of traffic accidents. The utilization of such association in reducing such happenings of accidents is by providingprecise information regarding the current status of the road such as friction level to those who drive or to the smart intellectual system that devious the car. Algorithms meant for learning and searching endowed with the performance of various procedures, frameworks, nature and finally the people who opens the door to new possibilities of fundamental world based on the results of analyzing huge volume of data. In today’s fast changing world, it is a high demand in the need of high efficiency in transport industry as the traffic is keep increasing which may have a drastic effect on low margin business with a high turnover. Astonishing changes in exterior conditions like increase in fuel rates, economic down trends or vehicle failures may affect the profit or loss in businesses. This results in high need of continuous monitoring over the efficiency