Algoritmo De Treinamento Para Uma Rede SLFN Com Projeção Aleatória E Margem Larga Learning Algorithm for an SLFN Network with Random Projection and Large Margin

ABSTRACTThis work presents a large margin learning algorithm for singlehidden layer feedforward networks (SLFNs) with random weightsfor the hidden neurons, called RP-IMA. This algorithm, applied tobinary classification problems, proposes randomly assignedweightsto the hidden layer and the use of an incremental margin algorithmto calculate the weights of the output neuron of the SLFN. Theresults showed that the proposed algorithm is capable to obtaina large separation margin in the feature space and has its performanceless sensitive to variations in the network architecture, whencompared to Extreme Learning Machines.