Author: Piotr Sowiński
Keywords: logic programs, artificial neural networks, Python, neural-symbolic integration
In this work I describe process of implementation of an artificial neural network, which is a part of a neural-symbolic system used for abductive hypotheses generation. Implementation of the neural network and its algorithms has been made using Python–a high-level general-purpose programming language created in early 90′ by Guido van Rossum. In the paper I consider possible and adopted approaches and solutions in network architecture and its implementation. Afterwards, I also describe specific process of modeling an artificial neural network, its parameters and a learning algorithm.