References in periodicals archive ?
The highest RRs were achieved using MLP and PWWs of 32 ms and 128 ms, where a RR of 62 % was achieved.
Recognition rates for PWWs between 32 ms and 128 ms have risen on the average by 6.5 % in comparison with the results from Table I.
The MRA gained another 5.5 % using the narrower part and 6.7 % using the wider part of the PWWs on the average.
To obtain more detailed data, the size of the shift should be smaller than the processing window width (PWW).
The benefit of different features is dependent on the PWW, so using a fixed PWW presents a compromise with the features used.
With the use of Information Gain calculation we kept only the top 200 features for every PWW. In the second step the use of more sophisticated algorithms that evaluate different subsets is appropriate.
The removal of this limitation led to larger feature subsets that contained between 29 and 55 different features (variations depending on the PWW).
Walters is aware of the false universalism of the PWWS and the fact that that a person's participation and position in the highly stratified PLM has long been of central importance in defining the quality and extent of an their access to public provision.
This further diminishes the limited equality that the citizenship of the British PWWS promised.
This represents a significant qualitative shift away from the public welfare envisaged in the PWWS, built around notions of need and entitlement.
Acronyms browser ?
Full browser ?