A hierarchical framework is proposed to address the issues of modeling different type of words in keyword spotting (KWS). Keyword models are built at various levels according to the availability of training set resources for each individual word. The proposed approach improves the performance of KWS even when no training speech is available for the keywords. It also suggests an easier way to co...

In a recent paper, Dolado, Gonzalo and Mayoral (2002) introduce a fractional Dickey-Fuller (FD-F) t-statistic for testing a unit root against the alternative of a mean reverting fractional unit root process. This t-statistic is based on the assumption that the errors are unconditionally homoskedastic. However, Busetti and Taylor (2003), McConnell and Perez-Quiros (2000), and van Dijk et al. (20...

In this paper, we consider and examine the performance of two-step LM unit root tests with trend-breaks. In the first step, we jointly test for the existence and location of breaks using a maximum F-test. In the second step, we utilize the identified breaks and test for a unit root. A transformation procedure is adopted so that the tests with trend-breaks are invariant to nuisance parameters. W...

This paper proposes a simple and improved nonparametric unit–root test. An asymptotic distribution of the proposed test is established. Finite sample comparisons with an existing nonparametric test are discussed. Some issues about possible extensions are outlined.