— This paper presents a single-objective and a multiobjective stochastic optimization algorithms for global training of neural networks based on simulated annealing. The algorith...
This work provides an analysis of using the evolutionary algorithm EPNet to create ensembles of artificial neural networks to solve a range of forecasting tasks. Several previous...
Connectivity patterns in biological brains exhibit many repeating motifs. This repetition mirrors inherent geometric regularities in the physical world. For example, stimuli that ...
: A novel approach to artificial neural networks is presented. The philosophy of this approach is based on two aspects: the design of task-specific networks, and a new neuron model...
Benedito Dias Baptista F. Filho, Eduardo Lobo Lust...
By far, the support vector machines (SVM) achieve the state-of-theart performance for the text classification (TC) tasks. Due to the complexity of the TC problems, it becomes a ch...