Parallel factor (PARAFAC) analysis is an extension of a low rank decomposition to higher way arrays, usually called tensors. Most of existing methods are based on an alternating l...
Handling large amounts of data, such as large image databases, requires the use of approximate nearest neighbor search techniques. Recently, Hamming embedding methods such as spec...
Wireless sensor networks are deployed to monitor dynamic geographic phenomena, or objects, over space and time. This paper presents a new spatiotemporal data model for dynamic area...
Abstract. In this paper, we propose a method for blind source separation (BSS) of convolutive audio recordings with short blocks of stationary sources, i.e. dynamically changing so...
We discuss the problem of overfitting of probabilistic neural networks in the framework of statistical pattern recognition. The probabilistic approach to neural networks provides a...