Change Point Detection Approach Research Articles

Incipient fault detection in induction machine stator-winding using a fuzzy-Bayesian change point detection approach

In this paper the incipient fault detection problem in induction machine stator-winding is considered. The problem is solved using a new technique of change point detection in time series, based on a two-step formulation. The first step consists of a fuzzy clustering to transform the initial data, with arbitrary distribution, into a new one that can be approximated by a beta distribution. The fuzzy cluster centers are determined by using a Kohonen neural network. The second step consists in using the Metropolis–Hastings algorithm for performing the change point detection in the transformed time series generated by the first step with that known distribution. The incipient faults are detected as long as they characterize change points in such transformed time series. The main contribution of the proposed approach is the enhanced resilience of the new failure detection procedure against false alarms, combined with a good sensitivity that allows the detection of rather small fault signals. Simulation and practical results are presented to illustrate the proposed methodology.

Discussion on “Sequential Design and Estimation in Heteroscedastic Nonparametric Regression” by Sam Efromovich

The pros and cons of the wide class of regression functions are discussed, with a special concern about the preservation of essential characteristics. The practicality of the sequential approach for change-point detection is interesting. The merits of the asymptotic results for this should be explored. The possibility of using the present results for adaptive change-point detection should be observed.

A detection approach to search-space reduction for HMM state alignment in speaker verification

To support speaker verification (SV) in portable devices and in telephone servers with millions of users, a fast algorithm for hidden Markov model (HMM) alignment is necessary. Currently, the most popular algorithm is the Viterbi (1967) algorithm with beam search to reduce search-space; however, it is difficult to determine a suitable beam width beforehand. A small beam width may miss the optimal path while a large one may slow down the alignment. To address the problem, we propose a nonheuristic approach to reduce the search-space. Following the definition of the left-to-right HMM, we first detect the possible change-points between HMM states in a forward-and-backward scheme, then use the change-points to enclose a subspace for searching. The Viterbi algorithm or any other search algorithm can then be applied to the subspace to find the optimal state alignment. Compared to a full-search algorithm, the proposed algorithm is about four times faster, while the accuracy is still slightly better in an SV task; compared to the beam search algorithm, the proposed algorithm can provide better accuracy with even lower complexity. In short, for an HMM with S states, the computational complexity can be reduced up to a factor of S/3 with slightly better accuracy than in a full-search approach. This paper also discusses how to extend the change-point detection approach to large-vocabulary continuous speech recognition.