Bounded error parameter estimation: A sequential analytic center approach

In this paper, a sequential analytic center approach for bounded error para meter estimation is proposed. The authors show that the analytic center min imizes the logarithmic average output error among all the estimates within the membership set and is a maximum likelihood estimator for a class of noi se density functions which include parabolic densities and approximations o f truncated Gaussian. They also show that the analytic center is easily com putable for both offline and online problems with a sequential algorithm. T he convergence proof of this sequential algorithm is obtained and, moreover , it is shown that the complexity in terms of the maximum number of Newton iterations is linear in the number of observed data points.