Traffic Flow Density Estimation using Neural Networks

Estimating traffic flow densities measured through road sensors using neural networks

A parameter free method for estimating probability densities using a Neural Network is proposed. The method is applied to traffic flow data observed with road sensors is proposed. Estimating the joint distribution of adjacent sensors for the same time interval or for adjacent time intervals measured with the same sensor can be used to obtain more realistic imputations for missing or erroneous observations.