Pattern classification with Evolving Long-term Cognitive Networks

This paper presents an interpretable neural system—termed Evolving Long-term Cognitive Network—for pattern classification. The proposed model was inspired by Fuzzy Maps, which are recurrent networks for modeling and simulation. network architecture is comprised of two blocks: a input layer output layer. Network that gets unfolded in the same way as other networks, thus producing sort abstract hidden layers. In our model, we can attach meaningful linguistic labels to each neuron since neurons correspond features given classification problem class labels. Moreover, propose variant backpropagation learning algorithm compute required parameters. includes new regularization components aimed at obtaining more knowledge representations. numerical simulations using 58 datasets show achieves higher prediction rates when compared with traditional white boxes while remaining competitive black boxes. Finally, elaborate on interpretability system proof concept.