[hts-users:00220] Re: about delta
- Subject: [hts-users:00220] Re: about delta
- From: "Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxxx>
- Date: Wed, 29 Mar 2006 00:38:06 +0900
Hi,
liulei_198216@xxxxxxxxxxx wrote:
About the definition of Delta^(n) C_t ,
there are two parameters L(n) and w(i).
What is the meanings of them.
First of all, please read the paper carefully, I think definitions of all of them are written in the paper.
L_{-,+}(n) is the window length of the n-th dynamic feature and w^n(i) is the i-th regression window coefficient to
calculate the n-th dynamic feature.
For example,
\Delta c_t = 0.5*c_{t+1} - 0.5*c_{t-1} => L_{-}(1)=1, L_{+}(1)=1, w^{1}(-1)=-0.5, w^{1}(0)=0.0, w^{1}(1)=0.5
\Delta^2 c_t = c_{t+1} - 2*c_t + c_{t-1} => L_{-}(1)=1, L_{+}(1)=1, w^{1}(-1)=1.0, w^{1}(0)=-2.0, w^{1}(1)=1.0
Best regards,
Heiga Zen (Byung Ha Chun)
--
------------------------------------------------
Heiga ZEN (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)
Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan
http://kt-lab.ics.nitech.ac.jp/~zen
------------------------------------------------
- References
-
- [hts-users:00219] about delta, 刘 磊