Pub Date : 2006-12-01DOI: 10.1109/NSSPW.2006.4378837
Mo Chen, T. Gautama, D. Obradovic, J. Chambers, D. Mandic
A new framework for the assessment of the qualitative performance of Kalman filter is proposed. This is achieved by the recently proposed `Delay Vector Variance' (DVV) method for the signal modality characterisation, which is based upon the local predictability in the phase space. It is shown that Kalman filter not only outperforms common linear and non-linear filters in terms of quantitative performance but also achieves a better qualitative performance. A set of comprehensive simulations on representative data sets supports the analysis.
{"title":"Exploiting Signal Nongaussianity and Nonlinearity for Performance Assessment of Adaptive Filtering Algorithms: Qualitative Performance of Kalman Filter","authors":"Mo Chen, T. Gautama, D. Obradovic, J. Chambers, D. Mandic","doi":"10.1109/NSSPW.2006.4378837","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378837","url":null,"abstract":"A new framework for the assessment of the qualitative performance of Kalman filter is proposed. This is achieved by the recently proposed `Delay Vector Variance' (DVV) method for the signal modality characterisation, which is based upon the local predictability in the phase space. It is shown that Kalman filter not only outperforms common linear and non-linear filters in terms of quantitative performance but also achieves a better qualitative performance. A set of comprehensive simulations on representative data sets supports the analysis.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122738142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-13DOI: 10.1109/NSSPW.2006.4378813
S. Saha, P. Mandal, Y. Boers, H. Driessen
In this article we introduce a new proposal distribution to be used in conjunction with the sequential Monte Carlo (SMC) method of solving non-linear filtering problem. The proposal distribution incorporates all the information about the to be estimated current state form both the available state and observation processes. This makes it more effective than the state transition density which is more commonly used but ignores the recent observation. Because of its Gaussian nature it is also very easy to implement. We show further that the introduced proposal performs better than other similar importance functions which also incorporate both state and observations.
{"title":"Exact Moment Matching for Efficient Importance Functions in SMC Methods","authors":"S. Saha, P. Mandal, Y. Boers, H. Driessen","doi":"10.1109/NSSPW.2006.4378813","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378813","url":null,"abstract":"In this article we introduce a new proposal distribution to be used in conjunction with the sequential Monte Carlo (SMC) method of solving non-linear filtering problem. The proposal distribution incorporates all the information about the to be estimated current state form both the available state and observation processes. This makes it more effective than the state transition density which is more commonly used but ignores the recent observation. Because of its Gaussian nature it is also very easy to implement. We show further that the introduced proposal performs better than other similar importance functions which also incorporate both state and observations.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132455596","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378839
P. Fearnhead, O. Papaspiliopoulos, G. Roberts
In this short communication we present our recent work on the construction of novel particle filters for a class of partially-observed continuous-time dynamic models where the signal is given by a multivariate diffusion process; details are deferred to [1]. Our approach directly covers a variety of observation schemes, including diffusion observed with error, observation of a subset of the components of the multivariate diffusion and arrival times of a Poisson process whose intensity is a known function of the diffusion (Cox process). Unlike available methods, our particle filters do not require approximations of the transition and/or the observation density using time-discretisations. Instead, they build on recent methodology for the exact simulation of diffusion process and the unbiased estimation of the transition density as described in the recent article [2]. In particular, we require the Generalised Poisson Estimator, which is developed in [1]. Thus, our filters avoid the systematic biases caused by time-discretisations and they have significant computational advantages over alternative continuous-time filters. These advantages are supported by a central limit theorem.
{"title":"Particle Filtering for Diffusions Avoiding Time-Discretisations","authors":"P. Fearnhead, O. Papaspiliopoulos, G. Roberts","doi":"10.1109/NSSPW.2006.4378839","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378839","url":null,"abstract":"In this short communication we present our recent work on the construction of novel particle filters for a class of partially-observed continuous-time dynamic models where the signal is given by a multivariate diffusion process; details are deferred to [1]. Our approach directly covers a variety of observation schemes, including diffusion observed with error, observation of a subset of the components of the multivariate diffusion and arrival times of a Poisson process whose intensity is a known function of the diffusion (Cox process). Unlike available methods, our particle filters do not require approximations of the transition and/or the observation density using time-discretisations. Instead, they build on recent methodology for the exact simulation of diffusion process and the unbiased estimation of the transition density as described in the recent article [2]. In particular, we require the Generalised Poisson Estimator, which is developed in [1]. Thus, our filters avoid the systematic biases caused by time-discretisations and they have significant computational advantages over alternative continuous-time filters. These advantages are supported by a central limit theorem.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127014905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378823
D. Crisan
I report on a new class of algorithms for the numerical solution of the continuous time filtering problem. These algorithms are inspired by recent advances in the area of weak approximations for solutions of stochastic differential equations. The algorithms belonging to this class generate approximations of the conditional distribution of the signal in the form of linear combinations of Dirac measures, hence can be interpreted as particle filters or, more precisely, particle approximations to the solution of the filtering problem. The main characteristics of these algorithms are discussed and a convergence result for the entire class is stated.
{"title":"Particle Filters in a Continuous Time Framework","authors":"D. Crisan","doi":"10.1109/NSSPW.2006.4378823","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378823","url":null,"abstract":"I report on a new class of algorithms for the numerical solution of the continuous time filtering problem. These algorithms are inspired by recent advances in the area of weak approximations for solutions of stochastic differential equations. The algorithms belonging to this class generate approximations of the conditional distribution of the signal in the form of linear combinations of Dirac measures, hence can be interpreted as particle filters or, more precisely, particle approximations to the solution of the filtering problem. The main characteristics of these algorithms are discussed and a convergence result for the entire class is stated.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123006418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378827
A. Brockwell
We discuss the problem of "decoding" intended hand motion from direct measurement of neurons in the motor cortex, for the purpose of driving a prosthetic device. By building probabilistic models and making use of nonlinear non-Gaussian filtering techniques, we are able to obtain estimates of intended hand motion, along with associated standard errors. We use a refinement of a previous state-of-the-art model, and demonstrate how the filtering approach works in analysis of multi-neuron recordings collected from a monkey carrying out a "center-out" task.
{"title":"Filtering of Neural Signals for Mental Control of Robotic Prosthetic Devices","authors":"A. Brockwell","doi":"10.1109/NSSPW.2006.4378827","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378827","url":null,"abstract":"We discuss the problem of \"decoding\" intended hand motion from direct measurement of neurons in the motor cortex, for the purpose of driving a prosthetic device. By building probabilistic models and making use of nonlinear non-Gaussian filtering techniques, we are able to obtain estimates of intended hand motion, along with associated standard errors. We use a refinement of a previous state-of-the-art model, and demonstrate how the filtering approach works in analysis of multi-neuron recordings collected from a monkey carrying out a \"center-out\" task.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129752169","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378818
S. Maskell, Ben Alun-Jones, M. Macleod
Particle filters are often claimed to be readily parallelisable. However, the resampling step is non-trivial to implement in a fine-grained parallel architecture. While approaches have been proposed that modify the particle filter to be amenable to such implementation, this paper's novelty lies in its description of a Single Instruction Multiple Data (SIMD) implementation of a particle filter that uses N processors to process N particles. The resulting algorithm has a time complexity of O((log N)2) when performing resampling using N processors. The algorithm has been implemented using C for Graphics (CG), a language that enables the heavily pipelined architecture of modern graphics cards to be used to imitate a SIMD processor. Initial results are presented.
粒子滤波通常被认为是易于并行的。但是，在细粒度并行体系结构中实现重新采样步骤是非常重要的。虽然已经提出了修改粒子滤波器以适应这种实现的方法，但本文的新颖之处在于它描述了使用N个处理器处理N个粒子的粒子滤波器的单指令多数据(SIMD)实现。当使用N个处理器进行重采样时，所得算法的时间复杂度为O((log N)2)。该算法是使用C for Graphics (CG)实现的，这种语言使现代显卡的大量流水线架构能够用来模拟SIMD处理器。提出了初步结果。
{"title":"A Single Instruction Multiple Data Particle Filter","authors":"S. Maskell, Ben Alun-Jones, M. Macleod","doi":"10.1109/NSSPW.2006.4378818","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378818","url":null,"abstract":"Particle filters are often claimed to be readily parallelisable. However, the resampling step is non-trivial to implement in a fine-grained parallel architecture. While approaches have been proposed that modify the particle filter to be amenable to such implementation, this paper's novelty lies in its description of a Single Instruction Multiple Data (SIMD) implementation of a particle filter that uses N processors to process N particles. The resulting algorithm has a time complexity of O((log N)2) when performing resampling using N processors. The algorithm has been implemented using C for Graphics (CG), a language that enables the heavily pipelined architecture of modern graphics cards to be used to imitate a SIMD processor. Initial results are presented.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114082626","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378834
James Loxam, T. Drummond
Filtering is a key component of many modem control systems: from noisy measurements, we want to be able to determine the state of some system as it evolves over time. Modem applications that require filtering tend to implement a filter from one of two main families of techniques: the Kalman filter (and associated extensions) and the particle filter. Each is popular and correct in its own right for certain applications, however each also has its limitations making it unsuitable for other applications. In this paper we propose a new filter based on the Student-t distribution to address the problems of the aforementioned filters: a filter which admits multimodal state hypotheses, is more robust to outliers, and remains computationally tractable in high-dimensional spaces.
{"title":"Efficient Parametric Non-Gaussian Dynamical Filtering","authors":"James Loxam, T. Drummond","doi":"10.1109/NSSPW.2006.4378834","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378834","url":null,"abstract":"Filtering is a key component of many modem control systems: from noisy measurements, we want to be able to determine the state of some system as it evolves over time. Modem applications that require filtering tend to implement a filter from one of two main families of techniques: the Kalman filter (and associated extensions) and the particle filter. Each is popular and correct in its own right for certain applications, however each also has its limitations making it unsuitable for other applications. In this paper we propose a new filter based on the Student-t distribution to address the problems of the aforementioned filters: a filter which admits multimodal state hypotheses, is more robust to outliers, and remains computationally tractable in high-dimensional spaces.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124483691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378812
Niilo Sirola, S. Ali-Loytty, R. Piché
Algorithm developers need relevant and practical criteria to evaluate and compare the performance of different discrete-time filters or filter variants. This paper discusses some pit-falls in different approaches and proposes a combination of criteria on which to base comparisons. A comparison of eight filters for a class of hybrid personal positioning problems is presented as an example.
{"title":"Benchmarking Nonlinear Filters","authors":"Niilo Sirola, S. Ali-Loytty, R. Piché","doi":"10.1109/NSSPW.2006.4378812","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378812","url":null,"abstract":"Algorithm developers need relevant and practical criteria to evaluate and compare the performance of different discrete-time filters or filter variants. This paper discusses some pit-falls in different approaches and proposes a combination of criteria on which to base comparisons. A comparison of eight filters for a class of hybrid personal positioning problems is presented as an example.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"148 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134022967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378848
A. Lecchini, W. Glover, J. Lygeros, Jan Maciejowski
Markov chain Monte Carlo (MCMC) methods can be used to make optimal decisions in very complex situations in which stochastic effects are prominent. In this paper we briefly introduce our current research on the application of MCMC to the predictive control of complex stochastic systems and the application to air traffic control.
{"title":"Predictive Control of Complex Stochastic Systems using Markov Chain Monte Carlo with Application to Air Traffic Control","authors":"A. Lecchini, W. Glover, J. Lygeros, Jan Maciejowski","doi":"10.1109/NSSPW.2006.4378848","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378848","url":null,"abstract":"Markov chain Monte Carlo (MCMC) methods can be used to make optimal decisions in very complex situations in which stochastic effects are prominent. In this paper we briefly introduce our current research on the application of MCMC to the predictive control of complex stochastic systems and the application to air traffic control.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"149 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124152268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2006-09-01DOI: 10.1109/NSSPW.2006.4378852
D. Myatt, S. Nasuto, S. Maybank
The 3D reconstruction of a Golgi-stained dendritic tree from a serial stack of images captured with a transmitted light bright-field microscope is investigated. Modifications to the boot-strap filter are discussed such that the tree structure may be estimated recursively as a series of connected segments. The tracking performance of the bootstrap particle filter is compared against Differential Evolution, an evolutionary global optimisation method, both in terms of robustness and accuracy. It is found that the particle filtering approach is significantly more robust and accurate for the data considered.
{"title":"Towards the Automatic Reconstruction of Dendritic Trees using Particle Filters","authors":"D. Myatt, S. Nasuto, S. Maybank","doi":"10.1109/NSSPW.2006.4378852","DOIUrl":"https://doi.org/10.1109/NSSPW.2006.4378852","url":null,"abstract":"The 3D reconstruction of a Golgi-stained dendritic tree from a serial stack of images captured with a transmitted light bright-field microscope is investigated. Modifications to the boot-strap filter are discussed such that the tree structure may be estimated recursively as a series of connected segments. The tracking performance of the bootstrap particle filter is compared against Differential Evolution, an evolutionary global optimisation method, both in terms of robustness and accuracy. It is found that the particle filtering approach is significantly more robust and accurate for the data considered.","PeriodicalId":388611,"journal":{"name":"2006 IEEE Nonlinear Statistical Signal Processing Workshop","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2006-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127125716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：