The quality and precision of tracking maneuvering targets under heavy clutter is highly dependent on both the data association and the state estimation algorithms. In this study, measurement-to-track association problem for a single target when P-D = 1 is discussed. The problem considers the batch set of measurements in a time interval. An approximate stochastic optimization algorithm for data association is presented. To reduce the computational load, the rollout algorithm is utilized. The algorithm is applied to a tracking scenario and GNN, JPDA and MHT algorithms are compared with their rollout versions. In the comparison several different track quality measures are used to demonstrate the efficiency of the algorithm.