ABSTRACT
This chapter discusses several neuron models based on factor space. Factor space offers a mathematical frame of describing objects and concepts. The chapter introduces several factor space canes and switch factors and their growth relation. It explains class partition for multifactorial fuzzy decision-making using factor space canes. A factor space can be regarded as a “transformer”. It is well known that the plasticity of connection weights is a key basic of the learning in the neural network. A neuron is basically made up of five parts: cell body, cell membrane, dendrite, axon, and synapse. An axon can transmit the electric impulse signals in its cell body to other neurons through its synapses. Dendrites receive the electric signals from other neurons. The basic form of a neuron connecting with other neurons is relatively steady such that the electric signals which flow into its dendrites should classify the whole “input channels” into two classes: one is excitatory and the other is inhibitory.
