Add to Chrome
Add to Firefox
Add to EdgeNucleus Neural Network for Super Robust Learning
Apr 08, 2019
Artificial neural networks which model the neurons and connecting architectures in brain have achieved great successes in many problems, especially those with deep layers. In this paper, we propose a nucleus neural network (NNN) and corresponding architecture and parameter learning methods. In a nucleus, there are no regular layers, i.e., a neuron may connect to all the neurons in the nucleus. This architecture gets rid of layer limitation and may lead to more powerful learning capability. It is crucial to determine the connections given numerous neurons. Based on the principle that more relevant input and output neuron pair deserves higher connecting density, we propose an architecture learning model for the nucleus. Moreover, we propose an improved learning method for learning connecting weights and biases with the optimized architecture. We find that this novel architecture is robust to irrelevant components in test data. So we define a super robust learning problem and test the proposed network with one case where the types of image backgrounds in training and test sets are different. Experiments demonstrate that the proposed learner achieves significant improvement over traditional learners on the reconstructed data set.
