Knowledge-driven Board-Level Functional Fault Diagnosis
Introduction to Manufacturing TestSystem and TestsTesting in the Manufacturing LineIntroduction to Board-Level DiagnosisReview of State-of-the-ArtRule-Based MethodsModel-Based MethodsReasoning-Based MethodsAutomation in Diagnosis SystemNew Directions Enabled by Machine LearningResearch on Increasing Diagnosis AccuracyOther Relevant ResearchChallenges and OpportunitiesCost of Manufacturing TestReduction of Yield LossAccommodate New Assembly TechnologyExpand Industry Adoption and UseOutline of BookReferencesDiagnosis Using Support Vector Machines (SVM)Background and Chapter HighlightsDiagnosis Using Support Vector MachinesSupport Vector MachinesDemonstration of SVM-Based Diagnosis SystemSVM Diagnosis FlowMulti-kernel Support Vector Machines and Incremental LearningMulti-kernel Support Vector MachinesKernelMulti-kernel Support Vector MachinesIncremental LearningResultsEvaluation of MK-SVM-Based Diagnosis SystemEvaluation of Incremental SVM-Based Diagnosis SystemEvaluation of Incremental MK-SVM-Based Diagnosis SystemSummaryReferencesDiagnosis Using Multiple Classifiers and Majority-Weighted Voting (WMV)Background and Chapter HighlightsArtificial Neural Networks (ANN)Architecture of ANNsDemonstration of ANN-Based Diagnosis SystemComparison Between ANNs and SVMsDiagnosis Using Weighted-Majority VotingWeighted-Majority VotingResultsEvaluation of ANNs-Based Diagnosis SystemEvaluation of SVMs-Based Diagnosis SystemEvaluation of WMV-Based Diagnosis SystemSummaryReferencesAdaptive Diagnosis Using Decision Trees (DT)Background and Chapter HighlightsDecision TreesTraining of Decision TreesInformation GainGini IndexExample of DT-Based Training and DiagnosisDiagnosis Using Incremental Decision TreesIncremental Tree NodeAddition of a CaseEnsuring the Best SplittingTree TranspositionDiagnosis Flow Based on Incremental Decision TreesResultsEvaluation ofDT-Based Diagnosis SystemEvaluation of Incremental DT-Based Diagnosis SystemSummaryReferencesInformation-Theoretic Syndrome and Root-Cause EvaluationBackground and Chapter HighlightsEvaluation Methods for Diagnosis SystemsSubset Selection for Syndromes AnalysisClass-Relevance StatisticsEvaluation and Enhancement FrameworkEvaluation and Enhancement ProcedureAn Example of the Proposed FrameworkResultsDemonstration of Syndrome AnalysisDemonstration of Root-Cause AnalysisSummaryReferencesHandling Missing SyndromesBackground and Chapter HighlightsMethods to Handle Missing SyndromesMissing-Syndrome-Tolerant Fault Diagnosis FlowMissing-Syndrome-Preprocessing MethodsComplete-Case AnalysisFractional InstancesLabel ImputationFeature SelectionComplete-Case AnalysisLabel ImputationResultsEvaluation of Label ImputationEvaluation of Feature Selection in Handling Missing SyndromesComparison of Different Missing-Syndrome Handling MethodsMissing-Syndrome Handling for SVM-Based Diagnosis SystemMissing-Syndrome Handling for ANN-Based Diagnosis SystemMissing-Syndrome Handling for NB-Based Diagnosis SystemMissing-Syndrome Handling for DT-Based Diagnosis SystemEvaluation of Training TimeSummaryReferencesKnowledge Discovery and Knowledge TransferBackground and Chapter HighlightsOverview of Knowledge Discovery and Transfer FrameworkKnowledge-Discovery MethodKnowledge-Transfer MethodResultsEvaluation of Knowledge-Transfer MethodEvaluation of Hybrid MethodSummaryReferencesDirections for Future Work
