The following information was submitted:
Transactions: INTERNATIONAL JOURNAL of CIRCUITS, SYSTEMS and SIGNAL PROCESSING
Transactions ID Number: 17-173
Full Name: Benabdellah Yagoubi
Position: Associate Professor
Age: ON
Sex: Male
Address: Faculty of science and engineering, electrical engineering, Mostaganem University, Algeria
Country: ALGERIA
Tel:
Tel prefix:
Fax:
E-mail address: sbmsvy@gmail.com
Other E-mails:
Title of the Paper: A relative geometric observation space for the linear modelling
Authors as they appear in the Paper: Benabdellah Yagoubi
Email addresses of all the authors: sbmsvy@gmail.com
Number of paper pages: 8
Abstract: In this work we consider the inputs as a base of a relative geometric linear space to observe the real model output. We use, therefore, a covariance concept by means of the input space metric tensor to determine, geometrically, the model output contra variant components which represent conventional linear model coefficients. We determine, in addition, the estimation error variance between the real model and the proposed model using a simple Pythagoras theorem. We show the simplicity of this method compared to that of the least mean square (LMS) and the Yule-Walker equations, in determining the Wiener filter, in estimating the autoregressive model coefficients and as well as in interpreting geometrically Kalman Algorithm.
Keywords: Kaman algorithm; Intrinsic behaviour; Metric tensor; Observation space; Linear model; Autoregressive model
EXTENSION of the file: .pdf
Special (Invited) Session: A geometric approach to the linear modelling
Organizer of the Session: 303-126
How Did you learn about congress:
IP ADDRESS: 41.110.149.131
