Improving Abductive Diagnosis Through Structural Features: A Meta-Approach

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Abstract

While abductive reasoning provides an intuitive approach
to diagnosis, its computational complexity remains an obstacle.
Even though certain model representations are tractable, computing
solutions for instances of reasonable size and complexity persists
to pose a challenge. Hence, the discovery of efficient methods
to derive abductive explanations presents itself as appealing research
area. In this paper, we investigate the structural properties inherent
to formalizations suitable for abductive failure localization. Based
on the features extracted we construct a meta-approach exploiting a
machine learning classifier to predict the abductive reasoning technique
yielding the “best” performance on a specific diagnosis scenario.
To assess whether the proposed attributes are in fact sufficient
for forecasting the appropriate abduction procedure and to evaluate
the efficiency of our algorithm selection in comparison to traditional
abductive reasoning approaches, we conducted an empirical experiment.
The results obtained indicate that the trained model is capable
of predicting the most efficient algorithm and further, we can show
that the meta-approach is capable of outperforming each single abductive
reasoning method investigated.
Originalspracheenglisch
TitelProceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16)
Herausgeber (Verlag)CEUR WS Proceedings
Seitenumfang9
BandVol-1626
PublikationsstatusVeröffentlicht - 13 Sep 2016

Fingerprint

Structural properties
Computational complexity
Classifiers
Experiments

Schlagwörter

    Dies zitieren

    Koitz, R., & Wotawa, F. (2016). Improving Abductive Diagnosis Through Structural Features: A Meta-Approach. in Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16) (Band Vol-1626). CEUR WS Proceedings.

    Improving Abductive Diagnosis Through Structural Features: A Meta-Approach. / Koitz, Roxane; Wotawa, Franz.

    Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16). Band Vol-1626 CEUR WS Proceedings, 2016.

    Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

    Koitz, R & Wotawa, F 2016, Improving Abductive Diagnosis Through Structural Features: A Meta-Approach. in Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16). Bd. Vol-1626, CEUR WS Proceedings.
    Koitz R, Wotawa F. Improving Abductive Diagnosis Through Structural Features: A Meta-Approach. in Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16). Band Vol-1626. CEUR WS Proceedings. 2016
    Koitz, Roxane ; Wotawa, Franz. / Improving Abductive Diagnosis Through Structural Features: A Meta-Approach. Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16). Band Vol-1626 CEUR WS Proceedings, 2016.
    @inproceedings{0a638f82d35d4a8b84d887f60e8643ed,
    title = "Improving Abductive Diagnosis Through Structural Features: A Meta-Approach",
    abstract = "While abductive reasoning provides an intuitive approachto diagnosis, its computational complexity remains an obstacle.Even though certain model representations are tractable, computingsolutions for instances of reasonable size and complexity persiststo pose a challenge. Hence, the discovery of efficient methodsto derive abductive explanations presents itself as appealing researcharea. In this paper, we investigate the structural properties inherentto formalizations suitable for abductive failure localization. Basedon the features extracted we construct a meta-approach exploiting amachine learning classifier to predict the abductive reasoning techniqueyielding the “best” performance on a specific diagnosis scenario.To assess whether the proposed attributes are in fact sufficientfor forecasting the appropriate abduction procedure and to evaluatethe efficiency of our algorithm selection in comparison to traditionalabductive reasoning approaches, we conducted an empirical experiment.The results obtained indicate that the trained model is capableof predicting the most efficient algorithm and further, we can showthat the meta-approach is capable of outperforming each single abductivereasoning method investigated.",
    keywords = "Abductive Diagnosis",
    author = "Roxane Koitz and Franz Wotawa",
    year = "2016",
    month = "9",
    day = "13",
    language = "English",
    volume = "Vol-1626",
    booktitle = "Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16)",
    publisher = "CEUR WS Proceedings",

    }

    TY - GEN

    T1 - Improving Abductive Diagnosis Through Structural Features: A Meta-Approach

    AU - Koitz, Roxane

    AU - Wotawa, Franz

    PY - 2016/9/13

    Y1 - 2016/9/13

    N2 - While abductive reasoning provides an intuitive approachto diagnosis, its computational complexity remains an obstacle.Even though certain model representations are tractable, computingsolutions for instances of reasonable size and complexity persiststo pose a challenge. Hence, the discovery of efficient methodsto derive abductive explanations presents itself as appealing researcharea. In this paper, we investigate the structural properties inherentto formalizations suitable for abductive failure localization. Basedon the features extracted we construct a meta-approach exploiting amachine learning classifier to predict the abductive reasoning techniqueyielding the “best” performance on a specific diagnosis scenario.To assess whether the proposed attributes are in fact sufficientfor forecasting the appropriate abduction procedure and to evaluatethe efficiency of our algorithm selection in comparison to traditionalabductive reasoning approaches, we conducted an empirical experiment.The results obtained indicate that the trained model is capableof predicting the most efficient algorithm and further, we can showthat the meta-approach is capable of outperforming each single abductivereasoning method investigated.

    AB - While abductive reasoning provides an intuitive approachto diagnosis, its computational complexity remains an obstacle.Even though certain model representations are tractable, computingsolutions for instances of reasonable size and complexity persiststo pose a challenge. Hence, the discovery of efficient methodsto derive abductive explanations presents itself as appealing researcharea. In this paper, we investigate the structural properties inherentto formalizations suitable for abductive failure localization. Basedon the features extracted we construct a meta-approach exploiting amachine learning classifier to predict the abductive reasoning techniqueyielding the “best” performance on a specific diagnosis scenario.To assess whether the proposed attributes are in fact sufficientfor forecasting the appropriate abduction procedure and to evaluatethe efficiency of our algorithm selection in comparison to traditionalabductive reasoning approaches, we conducted an empirical experiment.The results obtained indicate that the trained model is capableof predicting the most efficient algorithm and further, we can showthat the meta-approach is capable of outperforming each single abductivereasoning method investigated.

    KW - Abductive Diagnosis

    M3 - Conference contribution

    VL - Vol-1626

    BT - Proceedings of the International Workshop on Defeasible and Ampliative Reasoning (DARe-16)

    PB - CEUR WS Proceedings

    ER -