The key contribution of this thesis is introducing the dempstershafer theory of evidence as a generalizable sensor fusion solution to overcome the typical contextsensing difficulties, wherein some of the available information items are subjective, sensor observations probability objective chance distribution is not known accurately. This study proposes a method based on dempstershafer theory dst and fuzzy neural network fnn to improve the reliability of recognizing fatigue driving. Mar 22, 2012 the dempster shafer theory, a generalization of the bayesian theory, is based on the idea of belief and as such can handle ignorance. The initial work introducing dempster shafer ds theory is found in dempster 1967 and shafer 1976. Scalable clustering using the dempstershafer theory of evidence alireza chakeri.
Yet, most do not have a good way of dealing with ignorance. The following discussion within this section on dst has mostly been adopted from klir 11. Nsgaii is a multiobjective genetic algorithm developed by k. Data classification using the dempstershafer method. Ultimately, both can be used to combine new observations, with prior weights to develop new weights, f. The dempstershafer theory dst is a mathematical theory of evidence. Application of dempstershafer theory for fusion of lap. Dempstershafer ds theory is one of the main tools for reasoning about data obtained from multiple sources, subject to uncertain information 12, 16. A simple view of the dempstershafer theory of evidence. Fusion background the method outlined herein is developed based on the dempster shafer theory of evidence. The paper presents an algorithm for translating the subontologies containing con. Dempster shafer theory and statistical inference with weak.
Our approach is based on the dempstershafer theory that captures several characteristics of realworld applications and allows an efficient management of the uncertainty related to sensor readings. How is dempstershafer theory different from the bayesian. A method for recognizing fatigue driving based on dempster. This naturally leads us to choose the dempster shafer approach as our first sensor fusion implementation algorithm. Dempster shafer theory matlab free open source codes. Dempster shafer theory dst further generalizes this idea by. Introduction the ability to group complex data into a finite number of classes is important in data mining, and means that more useful decisions can be made based on the available information. This paper discusses the relationship between dempster shafer theory and the classical bayesian method, describes our sensor fusion research work using dempster shafer theory in comparison with the weighted sum of probability method.
Part of thecomputer sciences commons this thesis is brought to you for free and open access by the graduate school at scholar commons. The dempstershafer ds theory is an excellent method of information fusion. Learn more about machine learning, multiclass svm, dempster shafer theory. How to do decision fusion using dempster shafer theory for a. Theory a so appealing theory to represent uncertainty within our data fusion applications. Using the dempstershafer theory of evidence to resolve abox. It also mak es use the ds w eigh t con ict metric to prev en t the rob ot from acting on fault y observ. It is much newer theory, originating in the work of a. This paper discusses the relationship between dempster shafer theory and the classical bayesian method, describes our sensor fusion research work using dempster shafer theory in comparison with.
The dempstershafer theory of evidence is a formalism, which allows appropriate interpretation of extractors con. In the present work, a unified approach for robust deployment of fusionbased wsns is proposed. In proceedings of ieee international conference on ebusiness engineering pp. Mel siegel, chair daniel siewiorek jie yang wolfgang grimm, robert bosch corporation the robotics institute. Support for normalized as well as unnormalized belief functions different montecarlo algorithms for combining belief functions various methods related to the generalized bayesian theorem measures of uncertainty methods for constructing belief functions from. Dempster shafer ds theory serv es as the foundation for uncertain t y managemen t. Bayesian probability theory extends this idea and allows us to assign a likelihood of being true to every possible event. Sensor fusion for contextaware computing using dempster. The concept of multisensor information fusion model based on the dempstershafer theory is introduced. Emotion classifier based on evidence theory by dempstershafer.
The sfx hitecture is a generic sensor fusion system for autonomous mobile rob ots, suitable for a wide v ariet y of. Dempstershafer theory for sensor f usion in autonomous mobile rob ots robin r. The name ambiguity problem presents many challenges for scholar finding, citation analysis and other related research fields. The dempster shafer theory of evidence is a powerful method for combining measures of evidence from different classifiers.
The dempstershafer theory of evidence is a powerful method for combining measures of evidence from different classifiers. An alternative to the bayesian theory of evidence is dempstershafer theory which produces results similar to bayesian theory. The theory of belief functions, also referred to as evidence theory or dempstershafer theory dst, is a general framework for reasoning with uncertainty, with understood connections to other frameworks such as probability, possibility and imprecise probability theories. The dempstershafer theory, a generalization of the bayesian theory, is based on the idea of belief and as such can handle ignorance. An introduction to bayesian and dempstershafer data fusion.
In the absence of information, these methods must then make assumptions about the sensor data. The use of locally weighted regression for the data fusion. Dempster with extensions by glen shafer in 1987 shafer, 1976. Dempster shafertheory, on the other hand, handles this by assigning 0% belief to head and tail and assigning 100% belief to the set head, tail, meaning. Sensor fusion using dempstershafer theory of evidence in autonomous robot navigation 9 direction is also continuous is type 2. Dempstershafer theory, data classification, dempsters rule of combination. This paper presents a new classifier combination technique based on the dempster shafer theory of evidence. Dst originated when shafer extended the original work of dempster in 1976 10. Expert systemsdempstershafer theory wikibooks, open books. Dempstershafertheory dst further generalizes this idea by. Unsupervised author disambiguation using dempstershafer theory. In a previous paper 1, the authors have investigated dierent techniques to evalu. Faulttolerant interval estimation fusion by dempster. Because of the difference which is caused by the sensors, it is essential to deal with the.
Towards having computers understand human usersontext. In remote sensing applications, the expected number of classes may be large, in particular for agricultural areas. A dempstershafer method for multisensor fusion by bethany. In this section, we describe the background of our work, which includes an overview of the transferable belief model tbm smets and kennes, 1994, a subjective interpretation of the dempstershafer theory shafer, 1976. A special feature of this theory is the possibility to characterize uncer. This paper discusses the relationship between dempstershafer theory and the classical bayesian method, describes our sensor fusion research work using dempstershafer theory in comparison with the weighted sum of probability method. Sensor fusion using dempster shafer theory of evidence in autonomous robot navigation 9 direction is also continuous is type 2. A new technique for combining multiple classifiers using. Sensor fusion using dempstershafer theory for contextaware hci conference paper pdf available in conference record ieee instrumentation and measurement technology conference 1. A new multisensor information fusion model using dempster. An automatic water detection approach based on dempstershafer theory for multi spectral images. Hall, dempster shafer theory of evidence in single pass fuzzy c means, ieee international conference on fuzzy systems, pp. The dempster shafer theory is a mathematical theory of evidence based on belief functions and plausible reasoning, which is used to combine separate pieces of information evidence to calculate the probability of an event.
Dempstershafer theory of evidence2 dempstershafertheorydstheory. Dempstershafer theory for sensor f usion robin murph y. Dempstershafertheory, on the other hand, handles this by assigning 0% belief to head and tail and assigning 100% belief to the set head, tail, meaning. First, fnn is introduced to obtain the basic probability assignment bpa of each piece of evidence given the lack of a general solution to the. Follow 2 views last 30 days aparna raghunath on 27 may 2016.
We want to perform decision fusion of two multiclass svm systems using dempster shafer theory 0 comments. The success of a dsbased method depends on how the basic probability assignment bpa or probability mass function is defined. Dempstershafer theory is based on the nonclassical idea of mass as opposed to the wellunderstood probabilities of bayes theory. A library written in java for performing calculations in the dempstershafer theory of evidence. Scalable clustering using the dempstershafer theory of evidence. The resulting information of the architecture is combined using similar sensor subset and dissimilar sensor subset. Dempster in the context of statistical inference, the theory was later. This naturally leads us to choose the dempstershafer approach as our first sensor fusion implementation algorithm. Using the dempstershafer theory of evidence to resolve. Alina bruma, nicoleta iftimie, rozina steigmann, adriana savin, raimond grimberg subject. An automatic water detection approach based on dempster. Emotion classifier based on evidence theory by dempster shafer.
A curve has curvature continuity as well as tangent direction continuity is called type 3. Dempstershafer evidence theory of information fusion based on infoevolutionary value for ebusiness with continuous improvement. The concept of multisensor information fusion model based on the dempster shafer theory is introduced. The principal task of such reasoning is data fusion, or evidence accumulation.
Dempstershafer theory, with an example given of fusion using the cornerstone of the theory known as dempsters rule. A python library for performing calculations in the dempstershafer theory of evidence. Support for normalized as well as unnormalized belief functions different montecarlo algorithms for combining belief functions various methods related to the generalized bayesian theorem measures of uncertainty methods for constructing belief functions from data both python 2. Both the flexibility and the difficulty consist in defining the probability mass function. In this work we show that it can be viewed as a bayesian fusion algorithm and derive it using bayesian arguments. The dempstershafer theory the dempstershafer theory 8 allows us to combine evidence from different experts or other sources and provides a measure of con. Nguyen 1978 points out, soon after its introduction, t. Fusionbased surveillance wsn deployment using dempster. Scalable clustering using the dempstershafer theory of.
Eddy current and ultrasound data fusion using dempster. With formal logic, every event can be either true or false. To develop an adequate understanding of this theory re. Different than the bayes rule, dempstershafer theory represents evidence as possibilistic belief. Sensor data fusion for contextaware computing using. The paper concludes with a discussion of the results and in particular future directions of the work and how enhanced data sets could aid the evaluation of the approach. The foundations of ds have been laid out by dempster 24, 6 and shafer 18 22.
We present a dempstershafer evidence theory formulation of the engine fault diagnosis problem in terms of fault frame of discernment and mass functions, evidence combination, decision making rules, and fusion performance evaluation. Unsupervised author disambiguation using dempstershafer. For example, within the field of medical diagnosis, it is. In the absence of information, these methods must then make. This method measures driving states using multifeature fusion. S evidence generati on for dempster shafer fusion using. In this paper, we offer an unsupervised dempster shafer theory dst based hierarchical agglomerative clustering algorithm for author. Eddy current and ultrasound data fusion using dempster shafer theory author. Sensor data fusion for contextaware computing using dempster shafer theory huadong wu cmuritr0352 submitted in partial fulfillment of the requirements for the degree of doctor of philosophy in robotics thesis committee. To attack this issue, various disambiguation methods combined with separate disambiguation features have been put forward. Since its introduction the very name causes confusion, a more general term often used is belief functions both used intermittently here. A new technique for combining multiple classifiers using the. The initial work introducing dempstershafer ds theory is found in dempster 1967 and shafer 1976.
The dempstershafer ds theory provides an efficient framework to implement multisensor data fusion. How to do decision fusion using dempster shafer theory for. A lot of work has been done to reduce the complexity of computations used in information fusion with dempster s rule. We begin with an outline of bayes theory, using it to discuss wellknown quantities such as priors, likelihood and posteriors, and we provide the basic bayesian. When all of the required information is available, many data fusion methods provide a solid approach. Because of the difference which is caused by the sensors, it is essential to deal with the evidence with a method of. The tbm is a model for representing the quantified beliefs held by an agent at a given time on a given frame of discernment. The feasibility of this idea is demonstrated via a meetingparticipant.
In this paper, we offer an unsupervised dempstershafer theory dst based hierarchical agglomerative. The aim of this paper is to investigate how to use the contextual knowledge in order to improve the fusion process. This paper presents a new classifier combination technique based on the dempstershafer theory of evidence. Available evidence, even if incomplete, can be used to find imprecise probabilities for events. Faulttolerant interval estimation fusion by dempstershafer theory baohua li yunmin zhu. A verified matlab toolbox for the dempstershafer theory. How to do decision fusion using dempster shafer theory for a multiclass svm problem. A simple view of the dempstershafer theory of evidence and. We begin with an outline of bayes theory, using it to discuss wellknown quantities such as priors, likelihood and posteriors, and we provide the basic bayesian fusion equation. Consequently, we will present the ontology that we have built, called dsontology, according to the dempstershafer model, which helps to construct uncertain aspects into.
Fusion background the method outlined herein is developed based on the dempstershafer theory of evidence. In this work the dempstershafer ds theory has been used for fusing nondestructive inspection ndi data. Engine fault diagnosis based on multisensor information. Faulttolerant interval estimation fusion by dempstershafer. Pdf keypoint descriptor fusion with dempstershafer theory.
Situations and evidence for identity using dempstershafer. In presence of an overload of information and an unknown environment, the reliability of the sources of information or the sensors is usually unknown and thus cannot be used to re ne the fusion process. Mar 20, 2006 in this work the dempster shafer ds theory has been used for fusing nondestructive inspection ndi data. The dempster shafer ds theory is an excellent method of information fusion.
The fusion result is a discrete value or a label, which is determined by the corresponding maximum probability values. Prioritizing intrusion analysis using dempstershafer. Pdf decision fusion using fuzzy dempstershafer theory. A python library for performing calculations in the dempster shafer theory of evidence. In section 2, we introduce some preliminary concepts of the evidence theory. The sfx implemen tation of ds theory incorp orates evidence from sensor observ ations and domain kno wledge in to three lev els of p erceptual abstraction. Ngpm is the abbreviation of a nsgaii program in matlab, which is the implementation of nsgaii in matlab. The kalman filter is traditionally viewed as a predictioncorrection filtering algorithm. Sensor data fusion for contextaware computing using dempster.