Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
406279 | Neurocomputing | 2015 | 7 Pages |
This paper proposes a novel method for supervised subspace learning based on Single-hidden Layer Feedforward Neural networks. The proposed method calculates appropriate network target vectors by formulating a Bayesian model exploiting both the labeling information available for the training data and geometric properties of the training data, when represented in the feature space determined by the network׳s hidden layer outputs. After the calculation of the network target vectors, Extreme Learning Machine-based neural network training is applied and classification is performed using a Nearest Neighbor classifier. Experimental results on publicly available data sets show that the proposed approach consistently outperforms the standard ELM approach, as well as other standard methods.