Training the Stochastic Kinetic Model of Neuron for Calculation of an Object’s Position in Space

• DOI: 10.1007/s10846-019-01068-0
• Publication Date: 2019-07-31T09:02:50Z
• Authors: Aleksandra Świetl...; Krzysztof Kolanowski; Rafał Kapela
• Keywords: Position (finance); Biological neuron model; Hodgkin–Huxley model

In this paper we focus on the stochastic kinetic extension of well-known Hodgkin-Huxley model a biological neuron. We show gradient descent algorithm for training neuron model. comparison with use only three weights instead nine. that trained gives equally good results as model, while gain more concise mathematical description procedure. The is tested in solving problem approximation, where approximated function membrane potential obtained using different models was chosen. Additionally, present simple application, which connected outputs recurrent neural network form system, used to calculate Euler angles an object’s position space, based linear and angular acceleration, direction magnitude Earth’s magnetic field.