Where learning paths meet: Convergence and divergence of statistical and reinforcement learning | Synapse