Kirchhoff Institute for Physics, University of Heidelberg, Heidelberg, Germany.
Department of Physiology, University of Bern, Bern, Switzerland.
Sci Rep. 2018 Jul 13;8(1):10651. doi: 10.1038/s41598-018-28999-2.
Spiking networks that perform probabilistic inference have been proposed both as models of cortical computation and as candidates for solving problems in machine learning. However, the evidence for spike-based computation being in any way superior to non-spiking alternatives remains scarce. We propose that short-term synaptic plasticity can provide spiking networks with distinct computational advantages compared to their classical counterparts. When learning from high-dimensional, diverse datasets, deep attractors in the energy landscape often cause mixing problems to the sampling process. Classical algorithms solve this problem by employing various tempering techniques, which are both computationally demanding and require global state updates. We demonstrate how similar results can be achieved in spiking networks endowed with local short-term synaptic plasticity. Additionally, we discuss how these networks can even outperform tempering-based approaches when the training data is imbalanced. We thereby uncover a powerful computational property of the biologically inspired, local, spike-triggered synaptic dynamics based simply on a limited pool of synaptic resources, which enables them to deal with complex sensory data.
提出了执行概率推理的尖峰网络，既是皮质计算的模型，也是机器学习中解决问题的候选方案。然而，基于尖峰的计算在任何方面都优于非尖峰替代方案的证据仍然很少。我们提出，短期突触可塑性可以为尖峰网络提供与经典网络相比具有明显计算优势的方法。在从高维、多样化的数据集学习时，能量景观中的深度吸引子经常导致采样过程中的混合问题。经典算法通过采用各种回火技术来解决这个问题，这些技术既计算量大，又需要全局状态更新。我们展示了如何在具有局部短期突触可塑性的尖峰网络中实现类似的结果。此外，当训练数据不平衡时，我们还讨论了这些网络如何甚至可以超过基于回火的方法。因此，我们揭示了一种基于生物启发的、局部的、基于尖峰触发的突触动力学的强大计算特性，这种特性基于有限的突触资源池，使它们能够处理复杂的感官数据。
