Urbanisation and permission to own private vehicles promoted vehicular networks to be intelligent and automatic decision-makers. To reduce congestion on roads and provide services on time to the travellers, industry and academia need to work collectively to find optimal algorithms and solutions for Intelligent Transportation System (ITS). Cloud computing in the transportation domain plays a vital role in handling the vehicle’s big data. It allows people to think of autonomous driving, vehicle control and to design new intelligent systems. It is a promising technology for storing, communicating and computing vehicular data. Many researchers have proposed computation technologies to analyse the vehicular cloud network, urban traffic control, road safety measures, accidents alert passengers, object detection, cloud network security models, infotainment services, etc. From the perspective of transportation planning, big data has various challenges. Hence, there is a need for an algorithm that investigates its computability and computational complexity. Data is collected from multiple resources, sensors, roadside units, cameras, traffic signals, etc. The collected data may not contain all the desired attributes or not be structured. Hence data mining algorithms play a vital role to convert data into a structured format. Classification algorithms are performed on the structured data to identify the classes of datasets for analysis. To derive meaningful information from classified data, machine learning algorithms are applied. Machine learning algorithms are able to extract information automatically without any human assistance. Forthwith there is a paradigm shift in the vehicular cloud networks, which is acknowledged as edge computing existed at the end of the network. Edge nodes are proficient at storing, communicating, and computing vehicular data. This paradigm shift efficiently handles delay-sensitive applications, automatically disseminating emergency services information to travellers better than the cloud. However, edge nodes cannot store data for a longer period, so it is necessary to transfer the data to the cloud. To do so, the framework is designed and developed to prefetch information from vehicles to store at clouds and edge servers, process the data and utilise the processed data to inform drivers about road conditions. The expected results from the simulation are: to identify which technology is delay tolerance to the services delivered and identify which machine learning algorithm is perfectly suitable for handling big data and correct prediction. Simulation results show logistic regression is the most appropriate prediction algorithm with minimum classification and correct prediction. Later performance of cloud and edge are compared using delay, accuracy, throughput and execution time to show that edge computing is the most prominent technology in today’s world than cloud computing.