Q-AGREL: A biologically plausible reinforcement learning rule for image classification - 42Papers