ISSN : 3059-0604
We develop a classroom module that introduces the perceptron through the XOR gate, a compact example that links secondary-school mathematics (coordinate geometry and systems of inequalities) to the idea that composing simple linear separators can create non-linear decision boundaries. The module is organized as three lessons and includes activity outlines, student-facing proof sketches for the limitation of single-layer perceptrons on XOR, and a minimal two-hidden-unit construction that realizes XOR. An optional computational lab visualizes a learned decision boundary of a small multilayer perceptron trained on XOR.
