作者: Benjamin Torben-Nielsen , Klaus M. Stiefel
DOI: 10.1080/09548980902984833
关键词: Artificial neural network 、 Distribution function 、 Algorithm 、 Inverse 、 Neuron 、 Function space 、 Mathematics 、 Function (mathematics) 、 Models of neural computation 、 Free parameter
摘要: For many classes of neurons, the relationship between computational function and dendritic morphology remains unclear. To gain insights into this relationship, we utilize an inverse approach in which optimize model neurons with realistic morphologies ion channel distributions (of IKA ICaT) to perform a function. In study, desired is input-order detection: have respond differentially arrival two inputs different temporal order. There single free parameter function, namely, time lag arrivals inputs. Systematically varying allowed us map one axis space structure space. Because optimized known certainty, their thorough analysis provides neurons’ functions, morphologies, distributions, electrophysiological dynamics. Finally, discuss issues optimality nervous systems.