-
A Stochastically Evolving Non-local Search and Solutions to Inverse Problems with Sparse Data
Authors:
Mamatha Venugopal,
Ram Mohan Vasu,
Debasish Roy
Abstract:
Building upon our earlier work of a martingale approach to global optimization, a powerful stochastic search scheme for the global optimum of cost functions is proposed on the basis of change of measures on the states that evolve as diffusion processes and splitting of the state-space along the lines of a Bayesian game. To begin with, the efficacy of the optimizer, when contrasted with one of the…
▽ More
Building upon our earlier work of a martingale approach to global optimization, a powerful stochastic search scheme for the global optimum of cost functions is proposed on the basis of change of measures on the states that evolve as diffusion processes and splitting of the state-space along the lines of a Bayesian game. To begin with, the efficacy of the optimizer, when contrasted with one of the most efficient existing schemes, is assessed against a family of Np-hard benchmark problems. Then, using both simulated- and experimental data, potentialities of the new proposal are further explored in the context of an inverse problem of significance in medical imaging, wherein the superior reconstruction features of a global search vis-à-vis the commonly adopted local or quasi-local schemes are brought into relief.
△ Less
Submitted 22 December, 2015;
originally announced December 2015.
-
An ensemble Kushner-Stratonovich-Poisson filter for recursive estimation in nonlinear dynamical systems
Authors:
Mamatha Venugopal,
Ram Mohan Vasu,
Debasish Roy
Abstract:
Despite the numerous applications that may be expeditiously modelled by counting processes, stochastic filtering strategies involving Poisson-type observations still remain somewhat poorly developed. In this work, we propose a Monte Carlo stochastic filter for recursive estimation in the context of linear/nonlinear dynamical systems with Poisson-type measurements. A key aspect of the present devel…
▽ More
Despite the numerous applications that may be expeditiously modelled by counting processes, stochastic filtering strategies involving Poisson-type observations still remain somewhat poorly developed. In this work, we propose a Monte Carlo stochastic filter for recursive estimation in the context of linear/nonlinear dynamical systems with Poisson-type measurements. A key aspect of the present development is the filter-update scheme, derived from an ensemble approximation of the time-discretized nonlinear filtering equation, modified to account for Poisson-type measurements. Specifically, the additive update through a gain-like correction term, empirically approximated from the innovation integral in the filtering equation, eliminates the problem of particle collapse encountered in many conventional particle filters. Through a few numerical demonstrations, the versatility of the proposed filter is brought forth, first with application to filtering problems with diffusive or Poisson-type measurements and then to an automatic control problem wherein the extremization of the associated cost functional is achieved simply by an appropriate redefinition of the innovation process.
△ Less
Submitted 8 July, 2014;
originally announced July 2014.
-
A New Evolutionary Bayesian Approach Incorporating Additive Path Correction for Nonlinear Inverse Problems
Authors:
M Venugopal,
D Roy,
R M Vasu
Abstract:
An evolutionary form of a generalized Bayesian update method, which is strictly derivative- free yet directed through an additive update term based purely on the statistical moments of the design variables, is proposed for nonlinear inverse problems in general and applied in particular to an optical imaging problem, the ultrasound modulated optical tomography (UMOT). The additive update term, whic…
▽ More
An evolutionary form of a generalized Bayesian update method, which is strictly derivative- free yet directed through an additive update term based purely on the statistical moments of the design variables, is proposed for nonlinear inverse problems in general and applied in particular to an optical imaging problem, the ultrasound modulated optical tomography (UMOT). The additive update term, which bypasses most pitfalls of a conventional weight- based Bayesian update, results from a change of measures aimed at driving appropriately derived observation-prediction error terms or increments of cost functionals to zero-mean Brownian martingales. This constitutes a novel characterization corresponding to the extremization of the cost functional(s), where the design unknowns are represented as diffusion processes evolving with respect to a continuously parameterized iteration variable. This leads to a recursive prediction-update algorithm to implement the search. The scheme offers freedom from sample degeneracy and the accompanying divergence of the conventional weight-based Bayesian update schemes. We obtain the order of convergence of the conditioned process and also establish that the solutions are stable against tolerable variations in the regularizing noise terms, even as the original inverse problem remains severely ill-posed. Numerical evidence on solutions to the UMOT problem also confirms substantive improvements in the reconstruction efficacy through the proposed method vis-à- vis a Gauss-Newton approach, especially where the regularized quasi-Newton direction has low sensitivity to variations in the design unknowns.
△ Less
Submitted 16 December, 2013; v1 submitted 13 May, 2013;
originally announced May 2013.
-
A Kushner-Stratonovich Monte Carlo Filter Applied to Nonlinear Dynamical System Identification
Authors:
S Sarkar,
S R Chowdhury,
M Venugopal,
R M Vasu,
D Roy
Abstract:
A Monte Carlo filter, based on the idea of averaging over characteristics and fashioned after a particle-based time-discretized approximation to the Kushner-Stratonovich (KS) nonlinear filtering equation, is proposed. A key aspect of the new filter is the gain-like additive update, designed to approximate the innovation integral in the KS equation and implemented through an annealing-type iterativ…
▽ More
A Monte Carlo filter, based on the idea of averaging over characteristics and fashioned after a particle-based time-discretized approximation to the Kushner-Stratonovich (KS) nonlinear filtering equation, is proposed. A key aspect of the new filter is the gain-like additive update, designed to approximate the innovation integral in the KS equation and implemented through an annealing-type iterative procedure, which is aimed at rendering the innovation (observation-prediction mismatch) for a given time-step to a zero-mean Brownian increment corresponding to the measurement noise. This may be contrasted with the weight- based multiplicative updates in most particle filters that are known to precipitate the numerical problem of weight collapse within a finite-ensemble setting. A study to estimate the a-priori error bounds in the proposed scheme is undertaken. The numerical evidence, presently gathered from the assessed performance of the proposed and a few other competing filters on a class of nonlinear dynamic system identification and target tracking problems, is suggestive of the remarkably improved convergence and accuracy of the new filter.
△ Less
Submitted 16 December, 2013; v1 submitted 14 February, 2013;
originally announced February 2013.
