Perceptron
One of the earliest neural network models, proposed by Frank Rosenblatt. It is a single-layer linear classifier capable of handling only linearly separable problems and cannot solve nonlinear problems such as XOR. Its limitations drove the development of multi-layer neural networks.
Timeline (last 90 days)
Different machine learning architectures may assign different difficulty rankings to the same concept.
The XOR function is completely non-linearly separable for a single-layer linear model; this is the central counterexample put forward by Minsky and Papert in "Perceptrons" in 1969.
The XOR problem was a famous bottleneck historically encountered by the perceptron, directly driving the development of multi-layer networks and the backpropagation algorithm.
The perceptron convergence theorem proves that if the data is linearly separable, the algorithm is guaranteed to find the correct classification hyperplane within a finite number of steps
The XOR problem is the simplest linearly non-separable problem, and a single-layer perceptron cannot classify it correctly.
All Facts (6)
In 1969, Minsky and Papert proved that a single-layer perceptron cannot solve the XOR problem.
60%UnverifiedDifferent machine learning architectures may assign different difficulty rankings to the same concept.
50%UnverifiedThe XOR function is completely non-linearly separable for a single-layer linear model; this is the central counterexample put forward by Minsky and Papert in "Perceptrons" in 1969.
50%UnverifiedThe XOR problem was a famous bottleneck historically encountered by the perceptron, directly driving the development of multi-layer networks and the backpropagation algorithm.
50%UnverifiedThe perceptron convergence theorem proves that if the data is linearly separable, the algorithm is guaranteed to find the correct classification hyperplane within a finite number of steps
50%UnverifiedThe XOR problem is the simplest linearly non-separable problem, and a single-layer perceptron cannot classify it correctly.
50%