Parametric models allows learning by converging to the desired parameters, which are randomly initialized initially. Among the parametric models, especially in connectionist AI, neural networks are popular and widely used in literature.
In an artificial neural network, the building block is a neuron. Neuron consists of inputs, followed by weight connections, followed by summation unit, followed by non-linear activation function, followed by output.
Are there any parametric models in literature that are similar (and almost same) to artificial neural networks but differ only in basic building block?