The XOR problem involves the representation of a XOR gate in a neural network. This relatively simple modelling problem partially brought the whole study of neural networks to a standstill during the 1970s. The problem is that an elementary perceptron (one layer) has no hidden neurons and thus cannot classify linearly inseparable input patterns. Unfortunately, these patterns are common occurrences. The XOR gate is just such a problem and may be viewed as a special case of a more general problem. However, it was discovered that a two layer perceptron will successfully classify the input patterns.

Log in or register to write something here or to contact authors.