Traffic forecasting is very important in network flow to reduce the congestion and to get the best and short way data packets transmission. Also traffic flow is depends upon the chunk size and type whether it is image, audio, video, its format, size, resolution, used compression techniques, encryption, decryption header and algorithm information. Like this the traffic on the Indian roads are also a very huge and complicated issue and depends on multiple conditions like location, priority, weather, road condition, time , place etc. By using deep learning approaches accurate detection of traffic flow is possible. In this article we discuss the deep learning methodology which uses DBM approach to generate unsupervised learning datasets and RF approach to generate the exact and accurate traffic flow along with threshold based technique.

DBN, RF, DL,AI, Unsupervised learning.