This note develops robust filters for time-varying uncertain discrete-time systems. The developed filters are based on a data regularization solution and they enforce a minimum state-error variance property. Simulation results confirm their superior performance over other robust filter designs.