Kernel regression for travel time estimation via convex optimization

We develop an algorithm aimed at estimating travel time on segments of a road network using a convex optimization framework. Sampled travel time from probe vehicles are assumed to be known and serve as a training set for a machine learning algorithm to provide an optimal estimate of the travel time for all vehicles. A kernel method is introduced to allow for a non-linear relation between the known entry times and the travel times that we want to estimate. To improve the quality of the estimate we minimize the estimation error over a convex combination of known kernels. This problem is shown to be a semi-definite program. A rank-one decomposition is used to convert it to a linear program which can be solved efficiently.