Instinct and experience are shown to form a potent combination to achieve effective foraging in a simulated environment. A neural network capable of evolving instinct-related neur...
The use of monolithic neural networks (such as a multilayer perceptron) has some drawbacks: e.g. slow learning, weight coupling, the black box effect. These can be alleviated by t...
Frequently, sequences of state transitions are triggered by specific signals. Learning these triggered sequences with recurrent neural networks implies storing them as different at...
In this paper we present an intrusion detection engine comprised of two main elements; firstly, a neural network for the actual detection task and secondly watermarking techniques...
— The paper aims at designing a scheme for automatic identification of a species from its genome sequence. A set of 64 three-tuple keywords is first generated using the four type...