Growing Arti cial Neural Networks Basedon

| With this paper we propose a learning architecture for growing complex ar-tiicial neural networks. The complexity of the growing network is adapted automatically according to the complexity of the task. The algorithm generates a feed forward network bottom up by cyclically inserting cascaded hidden layers. Inputs of a hidden layer unit are locally restricted with respect to the input space by using a new kind of activation function. Contrary to the Cascade-Correlation Learning Architecture we introduce different correlation measures to train the network units featuring diierent goals. The task decomposition between subnetworks is done by maximizing the anticorrelation between the hidden layer units output and a connection routing algorithm between the hidden layers. These features resembles the TACO-MA (TAsk decomposition, COrrelation Measures and local Attention neurons) learning architecture. Results are shown for two dii-cult to solve problems in comparison to those produced by the CASCOR algorithm.