We propose a network architecture which uses a single internal layer of locally-tuned processing units to learn both classification tasks and real-valued function approximations (Moody and Darken 1988). We consider training such networks in a completely supervised manner, but abandon this approach in favor of a more computationally efficient hybrid learning method which combines self-organized and supervised learning. Our networks learn faster than backpropagation for two reasons: the local representations ensure that only a few units respond to any given input, thus reducing computational overhead, and the hybrid learning rules are linear rather than nonlinear, thus leading to faster convergence. Unlike many existing methods for data analysis, our network architecture and learning rules are truly adaptive and are thus appropriate for real-time use.

No abstract

We present and compare learning rate schedules for fast adaptive k-means clustering which surpass the standard MacQueen learning rate schedule in speed and quality of solution by several orders of magnitude for large k. Our methods accomplish this by largely overcoming the problems of metastable local minima and non-stationarity of cluster region boundaries which plague the MacQueen approach.

No abstract

In laser angioplasty, fluorescence spectra of targeted tissue may be used to classify the tissue as atherosclerotic or normal and guide selective laser ablation of atherosclerotic plaque. Here, the ability of the back-propagation and K-nearest neighbors techniques to classify arterial fluorescence spectra is investigated. Both methods are competitive with other classification schemes. The relative performance of variations on both techniques is used to make inferences about the geometry of the classification task.