A Biologically Motivated Synthesis Of Accumulator And Reinforcement-Learning Models For Describing Adaptive Decision-Making