Results 1  10
of
47
A backward particle interpretation of FeynmanKac formulae
 ESAIM: Math. Model. Num. Analy
, 2010
"... HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte p ..."
Abstract

Cited by 19 (7 self)
 Add to MetaCart
(Show Context)
HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et a ̀ la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. appor t d e r e cher c he
Recognizing recurrent neural networks (rrnn): Bayesian inference for recurrent neural networks. Biological cybernetics
, 2012
"... Recurrent neural networks (RNNs) are widely used in computational neuroscience and machine learning applications. In an RNN, each neuron computes its output as a nonlinear function of its integrated input. While the importance of RNNs, especially as models of brain processing, is undisputed, it ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
(Show Context)
Recurrent neural networks (RNNs) are widely used in computational neuroscience and machine learning applications. In an RNN, each neuron computes its output as a nonlinear function of its integrated input. While the importance of RNNs, especially as models of brain processing, is undisputed, it is also widely acknowledged that the computations in standard RNN models may be an oversimplification of what real neuronal networks compute. Here, we suggest that the RNN approach may be made both neurobiologically more plausible and computationally more powerful by its fusion with Bayesian inference techniques for nonlinear dynamical systems. In this scheme, we use an RNN as a generative model of dynamic input caused by the environment, e.g. of speech or kinematics. Given this generative RNN model, we derive Bayesian update equations that can decode its output. Critically, these updates define a ’recognizing RNN ’ (rRNN), in which neurons compute and exchange prediction and prediction error messages. The rRNN has several desirable features that a conventional RNN does not have, for example, fast decoding of dynamic stimuli and robustness to initial conditions and noise. Furthermore, it implements a predictive coding scheme for dynamic inputs. We suggest that the Bayesian inversion of recurrent neural networks may be useful both as a model of brain function and as a machine learning tool. We illustrate the use of the rRNN by an application to the online decoding (i.e. recognition) of human kinematics. 1
Identification of Mixed Linear/Nonlinear StateSpace Models
"... Abstract — The primary contribution of this paper is an algorithm capable of identifying parameters in certain mixed linear/nonlinear statespace models, containing conditionally linear Gaussian substructures. More specifically, we employ the standard maximum likelihood framework and derive an expec ..."
Abstract

Cited by 3 (2 self)
 Add to MetaCart
(Show Context)
Abstract — The primary contribution of this paper is an algorithm capable of identifying parameters in certain mixed linear/nonlinear statespace models, containing conditionally linear Gaussian substructures. More specifically, we employ the standard maximum likelihood framework and derive an expectation maximization type algorithm. This involves a nonlinear smoothing problem for the state variables, which for the conditionally linear Gaussian system can be efficiently solved using a so called RaoBlackwellized particle smoother (RBPS). As a secondary contribution of this paper we extend an existing RBPS to be able to handle the fully interconnected model under study. I.
An Online ExpectationMaximization Algorithm for Changepoint Models
"... Changepoint models are widely used to model the heterogeneity of sequential data. We present a novel sequential Monte Carlo (SMC) online ExpectationMaximization (EM) algorithm for estimating the static parameters of such models. The SMC online EM algorithm has a cost per time which is linear in th ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
(Show Context)
Changepoint models are widely used to model the heterogeneity of sequential data. We present a novel sequential Monte Carlo (SMC) online ExpectationMaximization (EM) algorithm for estimating the static parameters of such models. The SMC online EM algorithm has a cost per time which is linear in the number of particles and could be particularly important when the data is representable as a long sequence of observations, since it drastically reduces the computational requirements for implementation. We present an asymptotic analysis for the stability of the SMC estimates used in the online EM algorithm and demonstrate the performance of this scheme using both simulated and real data originating from DNA analysis. 1
Particle Filters for Joint Timing and Carrier Estimation: Improved Resampling Guidelines and Weighted Bayesian CramérRao Bounds
"... Abstract—This paper proposes a framework for joint blind timing and carrier offset estimation and data detection using a Sequential Importance Sampling (SIS) particle filter in Additive White Gaussian Noise (AWGN) channels. We assume baud rate sampling and model the intractable posterior probability ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
(Show Context)
Abstract—This paper proposes a framework for joint blind timing and carrier offset estimation and data detection using a Sequential Importance Sampling (SIS) particle filter in Additive White Gaussian Noise (AWGN) channels. We assume baud rate sampling and model the intractable posterior probability distribution functions for sampling timing and carrier offset particles using beta distributions. To enable the SIS approach to estimate static synchronization parameters, we propose new resampling guidelines for dealing with the degeneracy problem and fine tuning the estimated values. We derive the Weighted Bayesian Cramér Rao Bound (WBCRB) for joint timing and carrier offset estimation, which takes into account the prior distribution of the estimation parameters and is an accurate lower bound for all considered Signal to Noise Ratio (SNR) values. Simulation results are presented to corroborate that the Mean Square Error (MSE) performance of the proposed algorithm is close to optimal at higher SNR values (above 20 dB). In addition, the bit error rate performance approaches that of the perfectly synchronized case for small unknown carrier offsets and any unknown timing offset. The advantage of our particle filter algorithm, compared to existing techniques, is that it can work for the full range acquisition of carrier offsets. Index Terms—Synchronization, timing offsets, carrier offsets, particle filter, CramérRao bounds. I.
Nested particle filters for online parameter estimation in discretetime statespace Markov models
 ArXiv:1308.1883
, 2013
"... iv ..."
(Show Context)
Comparison of Stochastic Methods for Control in Air Traffic Management?
"... Abstract: This paper provides a direct comparison of two stochastic optimisation techniques (Markov Chain Monte Carlo and Sequential Monte Carlo) when applied to the problem of conflict resolution and aircraft trajectory control in air traffic management. The two methods are then also compared to an ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Abstract: This paper provides a direct comparison of two stochastic optimisation techniques (Markov Chain Monte Carlo and Sequential Monte Carlo) when applied to the problem of conflict resolution and aircraft trajectory control in air traffic management. The two methods are then also compared to another existing technique of MixedInteger Linear Programming which is also popular in distributed control.
Maximum likelihood estimation in mixed linear/nonlinear statespace models
 In Submitted to the 49th IEEE Conference on Decision and Control (CDC
, 2010
"... Technical reports from the Automatic Control group in Linköping are available from ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Technical reports from the Automatic Control group in Linköping are available from
IN GEODETIC AND GEOGRAPHIC INFORMATION TECHNOLOGIES
"... 2011 Approval of the thesis: ..."
(Show Context)