Introducing magnetic fields into artificial neural networks

In applying Maxwell´s field equations to an analysis of the Hodgkin-Huxley´s model, the authors observed the following physical phenomenon: that a fired neuron can yield magnetic fields. The magnetic signals from a neuron are important information, which may be supposed to be conscious control information. Therefore, a study of artificial neural networks should take this field effect into consideration. Accordingly, here a study on the behavior of a unit neuron in the field is made and a new neuron model proposed. This neuron model differs greatly from recent neural models in two of its properties: one property is the treatment of two distinct types of input-output information; and the other is that it is a three-dimensional model using field theory as its mathematical method. A mathematical memory-learning relation has been derived from these new neuron equations, and is called the Learning Rull. Two learning theorems are put forward, and the control mechanisms of memory are also discussed.