| Description |
Understanding how our behavior and perception are generated from the response of brain cells, called neurons, to various external (e.g., visual input) and internal (e.g., reward value) covariates is a central goal in systems and computational neuroscience. Neurons, as the basic computational units of our nervous system, communicate with each other by generating and transferring electric pulses, called action potentials, along the nerve fibers. The action potentials, referred to as spikes, are generated by the neurons as they receive chemical and electrical inputs from other cells [1]. The ability of a neuron to fire a spike, however, is affected by several extrinsic or intrinsic factors-some of which may not be even observable or measurable, resulting in an inherent variability in the recorded neuronal responses to particular input, meaning that a neuron will not fire the same sequence of spikes if the same stimulus is presented to it repeatedly. Therefore, the neuronal activity is generally modeled as a stochastic process, where the probability of an observed sequence of spiking events, termed spike train, is described as a function of the extrinsic events (such as the presented sensory stimulus) and the intrinsic factors (such as the state of the brain) [2]. Despite this variability in the neural response at a single cell level, the brain has a robust representation of the world. |
