Who proposed the first mathematical model of an artificial neuron in 1943?

Warren McCulloch and Walter Pitts proposed the first mathematical model of an artificial neuron in 1943. Their paper described how simplified neurons could perform logical operations using binary inputs and outputs.

The McCulloch–Pitts neuron represented a biological neuron as a threshold device. It added incoming signals, compared the result with a threshold, and produced an output when the threshold was reached. This abstraction helped connect neuroscience, mathematical logic, and early computing.

The model was not a learning algorithm and did not reproduce the full complexity of real nerve cells. Later systems, including the perceptron and multilayer neural networks, added adjustable parameters and learning procedures. Nevertheless, the 1943 paper became a foundational reference for connectionism and artificial neural-network research.