This dissertation treats of the problem of attributing the occurrence of an individual event or state to a single cause — a problem commonly understood either as a question of distinguishing the cause from the mere conditions or as a matter of singling out, from several causes, one cause, as the cause. The main purpose of the study is to clarify some basic concepts, and some criteria of ascertainment of the cause, that may be discerned in the literature on causal attribution. Special attention is devoted to how the adequacy of causal attributions depends on pragmatic factors. The study begins with an analysis of J. S. Mill’s distinction in A System of Logic between a scientific and a common-parlance approach to the problem of causal attribution. Mill’s assumption that causal attribution in science always requires a universal-law subsumption is then examined in the context of a general discussion of the range of applicability of the covering-law model of explanation. Mill’s scientific and common-parlance notions of cause are compared with R. G. Collingwood’s historical (sense-I) and scientific (sense-II and -III) notions of cause. It is argued that there are purposes of inquiry for which Mill’s common-parlance approach is more relevant to causal attribution in natural science than his scientific approach. And, more generally, it is argued that although law subsumptions are necessary for the ascertainment of the causes, more is often required for explaining the effect. Samuel Gorovitz’s differentiating-factor analysis is discussed, and limitations of the model are identified. The relevance of Morton White’s abnormalistic approach to historical research is also examined. Further, a number of objectivistic approaches are discussed, and it is argued that objectivity is not attainable in causal attributions in a sense in which it always implies an improvement of our ability to attribute the occurrence of an individual event or state to a single cause.

Decision-theoretic systems, such as Markov Decision Processes (MDPs), are used for sequential decision-making under uncertainty. MDPs provide a generic framework that can be applied in various domains to compute optimal policies. This thesis presents techniques that offer explanations of optimal policies for MDPs and then refine decision theoretic models (Bayesian networks and MDPs) based on feedback from experts. Explaining policies for sequential decision-making problems is difficult due to the presence of stochastic effects, multiple possibly competing objectives and long-range effects of actions. However, explanations are needed to assist experts in validating that the policy is correct and to help users in developing trust in the choices recommended by the policy. A set of domain-independent templates to justify a policy recommendation is presented along with a process to identify the minimum possible number of templates that need to be populated to completely justify the policy. The rejection of an explanation by a domain expert indicates a deficiency in the model which led to the generation of the rejected policy. Techniques to refine the model parameters such that the optimal policy calculated using the refined parameters would conform with the expert feedback are presented in this thesis. The expert feedback is translated into constraints on the model parameters that are used during refinement. These constraints are non-convex for both Bayesian networks and MDPs. For Bayesian networks, the refinement approach is based on Gibbs sampling and stochastic hill climbing, and it learns a model that obeys expert constraints. For MDPs, the parameter space is partitioned such that alternating linear optimization can be applied to learn model parameters that lead to a policy in accordance with expert feedback. In practice, the state space of MDPs can often be very large, which can be an issue for real-world problems. Factored MDPs are often used to deal with this issue. In Factored MDPs, state variables represent the state space and dynamic Bayesian networks model the transition functions. This helps to avoid the exponential growth in the state space associated with large and complex problems. The approaches for explanation and refinement presented in this thesis are also extended for the factored case to demonstrate their use in real-world applications. The domains of course advising to undergraduate students, assisted hand-washing for people with dementia and diagnostics for manufacturing are used to present empirical evaluations.

The writer has attempted to discuss the distinction between the necessary and the contingent. It begins with a criticism against the possibility for a so-called ‘a possible worlds realism’ to give a “philosophical explanation” of this distinction. The writer argues that this is impossible, since it requires that a notion of this distinction be already accepted (more precisely that the necessity of such a theory is already accepted). After this specific criticism, the writer intends to show that this is a more general problem that follows any explanation of the contingent/necessary distinction. The writer then discusses the counter-argument that the requirements placed on these explanations are set to high, therefore the writer shows in theory the problem can be solved and sketches a more specific way how to explain and show the basis for this distinction. / Författaren har avsett att diskutera distinktionen mellan det nödvändiga och det kontingent. Det börjar med en kritik mot möjligheten för en så kallad ’möjliga världars realism’ att ge en ”filosofisk förklaring” av denna distinktion. Författaren argumenterar för att detta är omöjligt, eftersom det kräver att en sådan distinktion redan är accepterad (mer specifikt att nödvändigheten av en sådan teori redan är accepterad). Efter denna specifika kriticism, så avser författaren visa att detta problem är generellt och att det följer alla försök att förklara den kontingenta/nödvändiga distinktionen. Författaren diskuterar sedan motargumentet att de krav som ställts på dessa förklaringar är för högt ställda, därför visar författaren hur problemet kan lösas i teorin och visar också en förenklad modell av en lösningsmetod.

Companies develop process models to explicitly describe their business operations. In the same time, business operations, business processes, must adhere to various types of compliance requirements. Regulations, e.g., Sarbanes Oxley Act of 2002, internal policies, best practices are just a few sources of compliance requirements. In some cases, non-adherence to compliance requirements makes the organization subject to legal punishment. In other cases, non-adherence to compliance leads to loss of competitive advantage and thus loss of market share. Unlike the classical domain-independent behavioral correctness of business processes, compliance requirements are domain-specific. Moreover, compliance requirements change over time. New requirements might appear due to change in laws and adoption of new policies. Compliance requirements are offered or enforced by different entities that have different objectives behind these requirements. Finally, compliance requirements might affect different aspects of business processes, e.g., control flow and data flow. As a result, it is infeasible to hard-code compliance checks in tools. Rather, a repeatable process of modeling compliance rules and checking them against business processes automatically is needed. This thesis provides a formal approach to support process design-time compliance checking. Using visual patterns, it is possible to model compliance requirements concerning control flow, data flow and conditional flow rules. Each pattern is mapped into a temporal logic formula. The thesis addresses the problem of consistency checking among various compliance requirements, as they might stem from divergent sources. Also, the thesis contributes to automatically check compliance requirements against process models using model checking. We show that extra domain knowledge, other than expressed in compliance rules, is needed to reach correct decisions. In case of violations, we are able to provide a useful feedback to the user. The feedback is in the form of parts of the process model whose execution causes the violation. In some cases, our approach is capable of providing automated remedy of the violation. / Firmen entwickeln Prozessmodelle um ihre Geschäftstätigkeit explizit zu beschreiben. Geschäftsprozesse müssen verschiedene Arten von Compliance-Anforderungen einhalten. Solche Compliance-Anforderungen entstammen einer Vielzahl von Quellen, z.B. Verordnung wie dem Sarbanes Oxley Act von 2002, interne Richtlinien und Best Practices. Die Nichteinhaltung von Compliance-Anforderungen kann zu gesetzlichen Strafen oder dem Verlust von Wettbewerbsvorteilen und somit dem Verlust von Marktanteilen führen. Im Gegensatz zum klassischen, domänen-unabhängigen Begriff der Korrektheit von Geschäftsprozessen, sind Compliance-Anforderungen domain-spezifisch und ändern sich im Laufe der Zeit. Neue Anforderungen resultieren aus neuen Gesetzen und der Einführung neuer Unternehmensrichtlinien. Aufgrund der Vielzahl der Quellen für Compliance-Anforderungen, können sie unterschiedliche Ziele verfolgen und somit widersprüchliche Aussagen treffen. Schließlich betreffen Compliance-Anforderungen verschiedene Aspekte von Geschäftsprozessen, wie Kontrollfluss- und Datenabhängigkeiten. Auf Grund dessen können Compliance-Prüfungen nicht direkt Hard-coded werden. Vielmehr ist ein Prozess der wiederholten Modellierung von Compliance-Regeln und ihrer anschließenden automatischen Prüfung gegen die Geschäftsprozesse nötig. Diese Dissertation stellt einen formalen Ansatz zur Überprüfung der Einhaltung von Compliance-Regeln während der Spezifikation von Geschäftsprozessen vor. Mit visuellen Mustern ist es möglich, Compliance-Regeln hinsichtlich Kontrollfluss- und Datenabhängigkeiten sowie bedingte Regeln zu spezifizieren. Jedes Muster wird in eine Formel der temporalen Logik abgebildet. Die Dissertation behandelt das Problem der Konsistenzprüfung zwischen verschiedenen Compliance-Anforderungen, wie sie sich aus unterschiedlichen Quellen ergeben können. Ebenfalls zeigt diese Dissertation, wie Compliance-Regeln gegen die Geschäftsprozesse automatisch mittels Model Checking geprüft werden. Es wird aufgezeigt, dass zusätzliche Domänen-Kenntnisse notwendig sind, um richtige Entscheidungen zu treffen. Der vorgestelle Ansatz ermöglicht nützliches Feedback für Modellierer im Fall eines Compliance-Verstoßes. Das Feedback wird in Form von Teilen des Prozessmodells gegeben, deren Ausführung die Verletzung verursacht. In einigen Fällen ist der vorgestellte Ansatz in der Lage, den Compliance-Verstoß automatisch zu beheben.

Spontanabort i tidig graviditet hos kvinnor är den vanligaste typen av spontanabort och drabbar 10-35 % av alla gravida kvinnor. Syftet var att beskriva kvinnors upplevelser av omvårdnad i samband med spontanabort i tidig graviditet. Metoden som användes var systematisk litteraturstudie som totalt inkluderade nio studier. Datainsamlingen genomfördes i databaserna Cinahl, PubMed och PsycINFO. Studierna kvalitetsgranskades, och resultatet analyserades utifrån syftet. Materialet färgkodades utifrån funna mönster och tre kategorier växte fram. Kategorierna blev slutligen: kvinnors upplevelser av emotionellt stöd, kvinnors upplevelser av information och förklaring samt kvinnors upplevelser och behov av eftervård. Litteraturstudiens övervägande resultat innefattade upplevelser av otillfredsställelse med omvårdnaden bland de drabbade kvinnorna. Många kvinnor upplevde ett osympatiskt bemötande, bristande medicinsk förklaring av orsaken till spontanaborten samt en avsaknad av eftervård. Slutsatsen var att majoriteten av kvinnorna inte upplevde tillfredsställelse med det emotionella stödet, den medicinska förklaringen samt att de saknade eftervård i samband med spontanaborten i tidig graviditet. Ny forskning föreslås ur patientperspektivet med fokus på kvinnornas unika känslor och behov och ur personalperspektivet om bemötande och eftervård av kvinnor som nyligen varit med om spontanabort i tidig graviditet. / Early pregnancy loss occurs to 10-35% of all pregnant women and it is the most common type of pregnancy loss. The aim was to describe women’s experiences of nursing care in relation to early pregnancy loss. The used method was systematic literature review and resulted in totally used nine articles. The data collection was accomplished in the databases Cinahl, PubMed and PsycINFO. The articles quality was checked and the result was analyzed with regard to the aim. The material was coded in to colours from the found patterns and three categories took place. The categories were finally women’s experiences of emotional support, women’s experiences of information and explanation and women’s experiences and needs of aftercare. The greater part of the result included experiences of dissatisfaction with the nursing care among the women. Many women experienced an unsympathetic attitude, lack of medical explanation to the course and lack of aftercare. The conclusion was that the majority of the women didn’t experience satisfaction with the emotional support, the medical explanation and the aftercare, in relation to early pregnancy loss. Suggestion for new research is made, with focus on women’s unique feelings and needs, and also with focus on the nurses’ attitude and aftercare of these women.

No description available.

Decision-theoretic systems, such as Markov Decision Processes (MDPs), are used for sequential decision-making under uncertainty. MDPs provide a generic framework that can be applied in various domains to compute optimal policies. This thesis presents techniques that offer explanations of optimal policies for MDPs and then refine decision theoretic models (Bayesian networks and MDPs) based on feedback from experts. Explaining policies for sequential decision-making problems is difficult due to the presence of stochastic effects, multiple possibly competing objectives and long-range effects of actions. However, explanations are needed to assist experts in validating that the policy is correct and to help users in developing trust in the choices recommended by the policy. A set of domain-independent templates to justify a policy recommendation is presented along with a process to identify the minimum possible number of templates that need to be populated to completely justify the policy. The rejection of an explanation by a domain expert indicates a deficiency in the model which led to the generation of the rejected policy. Techniques to refine the model parameters such that the optimal policy calculated using the refined parameters would conform with the expert feedback are presented in this thesis. The expert feedback is translated into constraints on the model parameters that are used during refinement. These constraints are non-convex for both Bayesian networks and MDPs. For Bayesian networks, the refinement approach is based on Gibbs sampling and stochastic hill climbing, and it learns a model that obeys expert constraints. For MDPs, the parameter space is partitioned such that alternating linear optimization can be applied to learn model parameters that lead to a policy in accordance with expert feedback. In practice, the state space of MDPs can often be very large, which can be an issue for real-world problems. Factored MDPs are often used to deal with this issue. In Factored MDPs, state variables represent the state space and dynamic Bayesian networks model the transition functions. This helps to avoid the exponential growth in the state space associated with large and complex problems. The approaches for explanation and refinement presented in this thesis are also extended for the factored case to demonstrate their use in real-world applications. The domains of course advising to undergraduate students, assisted hand-washing for people with dementia and diagnostics for manufacturing are used to present empirical evaluations.

Disertacija skirta geriausio paaiškinimo išvedimo (GPI) analizei. Pirma, tyrimas eksplikuoja, kas yra GPI, ką žymi sąvoka aiškinamoji galia. Antra, tyrimas įvertina GPI pretenzijų į tiesą apginamumą – ar teisinga yra hipotezė, jeigu ji yra geriausias tam tikro reiškinio paaiškinimas. GPI skleidžiamas kaip materialaus samprotavimo forma, kuri priskiria teisingumą toms hipotezėms, kurios tarp savo varžovių pasižymi aukščiausiu aiškinamųjų veiksnių laipsniu: mažiausiai prieštarauja pripažintam žinojimui, unifikuoja svarbiausius reiškinius, yra giliausias paaiškinimas ir yra paprasčiausias paaiškinimas. Kritikuojamas tikimybinis požiūris į GPI. Išskiriami ir aprašomi keturi literatūroje randami GPI kaip vedančio į tiesą pagrindimo būdai: patikimumo-koherentinis, evoliucinis, tikimybinis bei empirinis-istorinis. Teigiama, kad nors GPI padeda įvertinti paaiškinimų tikimybių pasiskirstymą ir yra plačiai paplitusi praktika, dėl prastos atrankos argumento, galimo aiškinamųjų veiksnių nebendramatiškumo, pesimistinės indukcijos argumento ir argumento „atsarga gėdos nedaro“ žinomi GPI kaip vedančio į tiesą pagrindimo būdai nėra pakankami – GPI išvadų teisingumas yra atsitiktinis. / The dissertation analyses inference to the best explanation (IBE). Firstly, the thesis explicates the concept of IBE, it analyzes what does the concept of explanatory power stands for. Secondly, the thesis evaluates the tenability of the truth aspirations of IBE, it analyzes to what extent a hypothesis can be claimed to be true, given that it is the best explanation for some phenomenon of interest. IBE is explicated as a form of material inference that ascribes truth to the hypothesis that has the highest degree of explanatory virtues among its competitors: is the most consistent with approved background knowledge, unifies the most the relevant phenomena, is the deepest explanation and is the most simple explanation. Probabilistic approaches to IBE are criticized. Four basic ways of justification of the truth-conduciveness of IBE that can be discerned in the literature are described: reliabilist-coherentist, evolutionary, probabilistic and empirical-historical. The thesis concludes that while IBE could facilitate the determination of probability distributions and is a wide psychological practice, due to the bad lot argument, possible incommensurability of explanatory virtues, pessimistic induction and better safe than sorry beliefs all the four ways of justification of IBE in terms of truth-conduciveness cannot be taken for granted which leaves IBE only accidentally valid.

Thèse numérisée par la Division de la gestion de documents et des archives de l'Université de Montréal

The dissertation analyses inference to the best explanation (IBE). Firstly, the thesis explicates the concept of IBE, it analyzes what does the concept of explanatory power stands for. Secondly, the thesis evaluates the tenability of the truth aspirations of IBE, it analyzes to what extent a hypothesis can be claimed to be true, given that it is the best explanation for some phenomenon of interest. IBE is explicated as a form of material inference that ascribes truth to the hypothesis that has the highest degree of explanatory virtues among its competitors: is the most consistent with approved background knowledge, unifies the most the relevant phenomena, is the deepest explanation and is the most simple explanation. Probabilistic approaches to IBE are criticized. Four basic ways of justification of the truth-conduciveness of IBE that can be discerned in the literature are described: reliabilist-coherentist, evolutionary, probabilistic and empirical-historical. The thesis concludes that while IBE could facilitate the determination of probability distributions and is a wide psychological practice, due to the bad lot argument, possible incommensurability of explanatory virtues, pessimistic induction and better safe than sorry beliefs all the four ways of justification of IBE in terms of truth-conduciveness cannot be taken for granted which leaves IBE only accidentally valid. / Disertacija skirta geriausio paaiškinimo išvedimo (GPI) analizei. Pirma, tyrimas eksplikuoja, kas yra GPI, ką žymi sąvoka aiškinamoji galia. Antra, tyrimas įvertina GPI pretenzijų į tiesą apginamumą – ar teisinga yra hipotezė, jeigu ji yra geriausias tam tikro reiškinio paaiškinimas. GPI skleidžiamas kaip materialaus samprotavimo forma, kuri priskiria teisingumą toms hipotezėms, kurios tarp savo varžovių pasižymi aukščiausiu aiškinamųjų veiksnių laipsniu: mažiausiai prieštarauja pripažintam žinojimui, unifikuoja svarbiausius reiškinius, yra giliausias paaiškinimas ir yra paprasčiausias paaiškinimas. Kritikuojamas tikimybinis požiūris į GPI. Išskiriami ir aprašomi keturi literatūroje randami GPI kaip vedančio į tiesą pagrindimo būdai: patikimumo-koherentinis, evoliucinis, tikimybinis bei empirinis-istorinis. Teigiama, kad nors GPI padeda įvertinti paaiškinimų tikimybių pasiskirstymą ir yra plačiai paplitusi praktika, dėl prastos atrankos argumento, galimo aiškinamųjų veiksnių nebendramatiškumo, pesimistinės indukcijos argumento ir argumento „atsarga gėdos nedaro“ žinomi GPI kaip vedančio į tiesą pagrindimo būdai nėra pakankami – GPI išvadų teisingumas yra atsitiktinis.