[http://citeseer.nj.nec.com/storkey98palimpsest.html]

AmosStorkey (1998)

"Palimpsest or ForgetfulLearningRules for attractor ["NeuralNetwork"]s do not suffer from catastrophic forgetting. Instead they selectively forget older memories in order to store new patterns. Standard palimpsest learning algorithms have a capacity of up to 0.05n, where n is the size of the network. Here a new learning rule is introduced. The rule is local and incremental. It is show that it has paimpsest properties, and it has a PalimpsestCapacity of about 0.25n, much higher than the capacity of standard palimpsest schemes. It is shown that the algorithm acts as an IteratedFunctionSequence on the space of matrices, and this is used to illustrate the performance of the LearningRule."