No abstract

In this paper, we analyse different techniques to detect the number of clusters in a dataset, also know as cluster validation techniques. We also propose a new algorithm based on the combination of several validation indexes to simultaneously validate several partitions of a dataset generated by different clustering techniques and object distances. The existing validation techniques as well as the combination algorithm have been tested on three data sets: a synthesized mixture of Gaussians data set, the NCI60 microarray data set, and the Iris data set. Evaluation results have shown the adequate performance of the proposed approach, even if the input validity scores fail to discover the true number of clusters.

Abstract-Current speech processing in cochlear implants use a filterbank to analyse audio signals into several frequency bands, each associated with one electrode. Because the processing is performed on input signal blocks of fixed sizes, the filterbank provides a unique time-frequency resolution to represent the various signal features. However, different components of audio signals may require different time-frequency resolutions for an accurate representation and perception. In this paper we investigate the influence on speech intelligibility in cochlear implants users when filterbanks with different time-frequency resolutions are used. In order to represent all signal features accurately, an adaptive filterbank has been developed that accepts input blocks of different sizes. The different resolutions required are achieved by adequately switching between block sizes depending on the input signal characteristics. The filterbank was incorporated into the commercial Advanced Combinational Encoder (ACE) and acutely tested on six cochlear implant recipients.

No abstract