Subject: Re: Convolutive feature extraction (Searching for papers) From: Paris Smaragdis <paris@xxxxxxxx> Date: Mon, 16 Oct 2006 22:29:19 -0400> You wrote in last sentence of discussion in this paper: > Unfortunately, what makes non-negativity constraints so > successfully, in many > domains, is still a subject under debate and investigation. > > Unfotunately. Really unfortunately? I would like to say > fortunately, instead. > Perhaps you meanwhile got aware of my paper Non-negativity makes sense when one casually talks about it, but it is a tricky notion to justify, and even more importantly analyze in mathematical terms. So it is unfortunate that one can't express in equations why things work as well as they do. If one isn't concerned with this particular pursuit then non-negativity can indeed be a very fortunate venture! Paris