Harris C M
Department of Ophthalmology, Great Ormond Street Hospital for Children NHS Trust, Institute of Child Health, University College London, UK.
J Neurosci Methods. 1998 Aug 31;83(1):73-88. doi: 10.1016/s0165-0270(98)00063-6.
Evolution is a closed stochastic optimisation process driven by the interaction between behaviour and environment towards local maxima in fitness. It is inferred that nervous systems are selected to provide optimal control of behaviour (the 'assumption of optimality'), such that for some behaviours, the expectation of future hazards to survival are minimised. This is illustrated by goal-directed saccades in which minimising total flight-time of primary and secondary movements provides a better fit to observations than simply minimising the error of the primary movement. This optimisation is extended to intra-movement trajectories, where low-bandwidth (smooth) velocity profiles provide a more satisfactory description of observations than simple bang-bang control. Since minimum-time behaviours cannot be controlled by error feedback, it is concluded that the cerebellum must be executing a real-time unreferenced optimisation process. This requires explorative as well as exploitative behaviour. Stochastic gradient descent is discussed as a possible means by which the cerebellum may optimise behaviour.
进化是一个封闭的随机优化过程，由行为与环境之间的相互作用驱动，朝着适应性的局部最大值发展。据推断，神经系统被选择来提供行为的最优控制（“最优性假设”），使得对于某些行为，未来生存风险的期望值被最小化。这在目标导向的扫视中得到了体现，其中将主要和次要运动的总飞行时间最小化比仅仅最小化主要运动的误差更符合观察结果。这种优化扩展到运动内部轨迹，其中低带宽（平滑）速度分布比简单的开关控制更能令人满意地描述观察结果。由于最短时间行为不能由误差反馈控制，因此得出结论，小脑必须执行实时无参考优化过程。这需要探索性以及利用性的行为。随机梯度下降被讨论为小脑可能优化行为的一种可能方式。
