Vehicular Movement Prediction Via Supervised Vector Machine

E. I. IHAMA; M. I. AKAZUE; V. A. AMENAGHAWON

E. I. IHAMA Edo State Polytechnic Usen
M. I. AKAZUE Delta State University, Abraka, Nigeria
V. A. AMENAGHAWON Igbinedion University

Keywords: Peak periods and gridlock, Vehicle prediction, Supervised Vector Machine, Traffic congestion

Abstract

In major urbanized cities, the increase in the number of vehicles on the road at
certain periods creates traffic gridlocks. Most road users at these peak periods
experience traffic congestion during these peak periods. This has resulted in the
loss of hours of work, delays in travel time, accidents, and even loss of life Most
special vehicles like ambulances and emergency vehicles are trapped in the grid
lock. Hence, there is a need to have a better predictive model for vehicle movement.
In this paper, a vehicle prediction system which was developed using Supervised
Vector Machine in a python environment using road variables, such as road
condition, type of vehicle, weather condition and time of day vehicle. The dataset
was obtained from Kaggle online dataset. It was evaluated using the following
evaluation metrics: Root Mean Square Error (RMSE 2.845), Sum of Square Error
(SSE 809.6798875823043), R-Square (R2 -0.14387416306270362), and Adjusted RSquare
(R2 -0.21767249616352324 ).

Author Biography

E. I. IHAMA, Edo State Polytechnic Usen

Department of Computer Science and Information Technology