A Note on Improved Loss Bounds for Multiple Kernel Learning | Read Paper on Bytez