Regularized Information MaximizationΒΆ