Optimal Error Bounds for Classification and Regression of Inhomogeneous Data in Wireless Networks
Under ergodic-type conditions on the data distributions, optimal error bounds can be achieved for classification and regression tasks using universal rules, with applications in wireless networks.