Jerome J. Braun

Dempster-Shafer theory and Bayesian reasoning in multisensor data fusion

Bayesian and Dempster-Shafer Theory based methods are among the alternative algorithmic approaches to multisensor data fusion. The two approaches differ significantly and the extent of their applicability to data fusion is still being debated. This paper presents a Monte Carlo simulation approach for a comparative analysis of a Dempster-Shafer Theory based on a Bayesian multisensor data fusion in the classification task domain, including the implementation of both formalisms, and the results of the Monte Carlo experiments of this analysis.

Hybrid methods for multisource information fusion and decision support

This paper presents the progress of an ongoing research effort in multisource information fusion for biodefense decision support. The effort concentrates on a novel machine-intelligence hybrid-of-hybrids decision support architecture termed FLASH (Fusion, Learning, Adaptive Super-Hybrid) we proposed. The highlights of FLASH discussed in the paper include its cognitive-processing orientation and the hybrid nature involving heterogeneous multiclassifier machine learning and approximate reasoning paradigms. Selected specifics of the FLASH internals, such as its feature selection techniques, supervised learning, clustering, recognition and reasoning methods, and their integration, are discussed. The results to date are presented, including the background type determination and bioattack detection computational experiments using data obtained with a multisensor fusion testbed we have also developed. The processing of imprecise information originating from sources other than sensors is considered. Finally, the paper discusses applicability of FLASH and its methods to complex battlespace management problems such as course-of-action decision support.

Computational intelligence in biological sensing

This paper presents an alternative, computational intelligence based paradigm for biological attack detection. Conventional approaches to this difficult problem include sensor technologies and analytical modeling approaches. However, the processes that constitute the environmental background as well as those which occur as the result of an attack are extremely complex. This phenomenological complexity, in terms of both physics and biology aspects, is a challenge difficult to overcome by conventional approaches. In contrast to such approaches, the proposed approach is centered on automatic learning to discriminate between sensor signals that are in a normal range from those that are likely to represent a biological attack. It is argued that constructing machine learning methods robust enough to perform such a task is often more feasible than constructing an adequate model that could form a basis for bioattack detection. The paper discusses machine learning and multisensor information fusion methods in the context of biological attack detection in a subway environment, including recognition architecture and its components. However, the applicability of the proposed approach is much broader than the subway bioattack protection case, extending to a wide range of CBR defense applications.

Sensor data fusion with support vector machine techniques

This paper presents an approach to multisensor data fusion based on the use of Support Vector Machines (SVM). The approach is investigated using simulated generic sensor data, representative of data imperfections that may be encountered in multisensor fusion applications. In particular the issue of data incompleteness is addressed and a method exploiting vicinity of training points is proposed for incompleteness correction. The paper also investigates applicability of vicinal kernels in SVM-based sensor data fusion.

Information fusion of a large number of sources with support vector machine techniques

Applications of information fusion include cases that involve a large number of information sources. Methods developed in the context of few information sources may not, and often do not, scale well to cases involving a large number of sources. This paper addresses specifically the problem of information fusion of large number of information sources. Performance of Support Vector Machine (SVM) based approach is investigated in input spaces consisting of thousands of information sources. Microarray pattern recognition, an important bioinformatics task with significant medical diagnostics applications, is considered from the information and sensor data fusion viewpoint, and recognition performance experiments conducted on microarray data are discussed. An approach involving high-dimensional input space partitioning is presented and its efficacy is investigated. The aspects of feature-level and decision-level fusion are discussed as well. The results indicate the feasibility of the SVM based information fusion with large number of information sources.

Information fusion and uncertainty management for biological multisensor systems

This paper investigates methods of decision-making from uncertain and disparate data. The need for such methods arises in those sensing application areas in which multiple and diverse sensing modalities are available, but the information provided can be imprecise or only indirectly related to the effects to be discerned. Biological sensing for biodefense is an important instance of such applications. Information fusion in that context is the focus of a research program now underway at MIT Lincoln Laboratory. The paper outlines a multi-level, multi-classifier recognition architecture developed within this program, and discusses its components. Information source uncertainty is quantified and exploited for improving the quality of data that constitute the input to the classification processes. Several methods of sensor uncertainty exploitation at the feature-level are proposed and their efficacy is investigated. Other aspects of the program are discussed as well. While the primary focus of the paper is on biodefense, the applicability of concepts and techniques presented here extends to other multisensor fusion application domains.

Multiclassifier information fusion methods for microarray pattern recognition

This paper addresses automatic recognition of microarray patterns, a capability that could have a major significance for medical diagnostics, enabling development of diagnostic tools for automatic discrimination of specific diseases. The paper presents multiclassifier information fusion methods for microarray pattern recognition. The input space partitioning approach based on fitness measures that constitute an a-priori gauging of classification efficacy for each subspace is investigated. Methods for generation of fitness measures, generation of input subspaces and their use in the multiclassifier fusion architecture are presented. In particular, two-level quantification of fitness that accounts for the quality of each subspace as well as the quality of individual neighborhoods within the subspace is described. Individual-subspace classifiers are Support Vector Machine based. The decision fusion stage fuses the information from mulitple SVMs along with the multi-level fitness information. Final decision fusion stage techniques, including weighted fusion as well as Dempster-Shafer theory based fusion are investigated. It should be noted that while the above methods are discussed in the context of microarray pattern recognition, they are applicable to a broader range of discrimination problems, in particular to problems involving a large number of information sources irreducible to a low-dimensional feature space.

Inner rehearsal modeling for cognitive robotics

This paper presents a biomimetic approach involving cognitive process modeling, for use in intelligent robot decisionmaking. The principle of inner rehearsal, a process believed to occur in human and animal cognition, involves internal rehearsing of actions prior to deciding on and executing an overt action, such as a motor action. The inner-rehearsal algorithmic approach we developed is posed and investigated in the context of a relatively complex cognitive task, an under-rubble search and rescue. The paper presents the approach developed, a synthetic environment which was also developed to enable its studies, and the results to date. The work reported here is part of a Cognitive Robotics effort in which we are currently engaged, focused on exploring techniques inspired by cognitive science and neuroscience insights, towards artificial cognition for robotics and autonomous systems.

Fusion of disparate information sources in a hybrid decision-support architecture

Disparity and uncertainty of information sources are both significant problems in information fusion. This paper investigates the problem of disparity in general, and in conjunction with FLASH - a hybrid information-fusion cognitive-processing approach we developed. Different forms of disparity are identified and their categorization is presented, and their implications on the information fusion processes are discussed. The issue of feature-level vs. decision-level fusion is investigated, and the methods of coping with disparity within FLASH are presented. Source uncertainty estimation techniques are discussed as well. Disparity studies and the results of computational experiments related to them are presented. These studies are suggestive of the potential of the FLASH hybrid approach for fusion of disparate information sources.

Multisensor data analysis and aerosol background characterization

A portable and extensible multisensor testbed for long-term multi-point aerosol background data collections has been developed. The primary objective of the testbed is to support investigations related to the information fusion, machine-intelligence based CB decision support architectrure, now under development at MIT Lincoln Laboratory. This paper describes major design features of the testbed and concentrates on the analysis and the results of multiple indoor data collections. Specifically, two deployments of the testbed for extensive indoor data collection campaigns are described. The indoor background characterization results are presented.