Counterfactual Explanations Under Model Multiplicity and Their Use in Computational Argumentation

Model Multiplicity (MM) arises when multiple, equally performing machine learning models can be trained to solve the same prediction task. Recent studies show that models obtained under MM may produce inconsistent predictions for the same input. When this occurs, it becomes challenging to provide counterfactual explanations (CEs), a common means for offering recourse recommendations to individuals negatively affected by models' predictions. In this paper, we formalise this problem, which we name recourse-aware ensembling, and identify several desirable properties which methods for solving it should satisfy. We show that existing ensembling methods, naturally extended in different ways to provide CEs, fail to satisfy these properties. We then introduce argumentative ensembling, deploying computational argumentation to guarantee robustness of CEs to MM, while also accommodating customisable user preferences. We show theoretically and experimentally that argumentative ensembling satisfies properties that the existing methods lack, and that the trade-offs are minimal wrt accuracy.

Explainability and trust are vital in cybersecurity operations reliant on AI for threat detection and response. Many AI models, especially deep learning ones, act as “black boxes,” limiting transparency and eroding trust. Explainable AI (XAI) addresses this by clarifying decisions through techniques like feature importance scoring and counterfactual explanations. XAI enhances trust, ensures compliance with regulations like GDPR, and strengthens human-AI collaboration. By fostering transparency, organizations can act swiftly on AI-driven alerts, improve operational efficiency, and establish ethical, resilient defenses against evolving cyber threats.

Explainable Artificial Intelligence (XAI) is a research area that clarifies AI decision-making processes to build user trust and promote responsible AI. Hence, a key scientific challenge in XAI is the development of methods that generate transparent and interpretable explanations while maintaining scalability and effectiveness in complex scenarios. Rule-based methods in XAI generate rules that can potentially explain AI inferences, yet they can also become convoluted in large scenarios, hindering their readability and scalability. Moreover, they often lack contrastive explanations, leaving users uncertain why specific predictions are preferred. To address this scientific problem, we explore the integration of computational argumentation—a sub-field of AI that models reasoning processes through defeasibility—into rule-based XAI systems. Computational argumentation enables arguments modelled from rules to be retracted based on new evidence. This makes it a promising approach to enhancing rule-based methods for creating more explainable AI systems. Nonetheless, research on their integration remains limited despite the appealing properties of rule-based systems and computational argumentation. Therefore, this study also addresses the applied challenge of implementing such an integration within practical AI tools. The study employs the Logic Learning Machine (LLM), a specific rule-extraction technique, and presents a modular design that integrates input rules into a structured argumentation framework using state-of-the-art computational argumentation methods. Experiments conducted on binary classification problems using various datasets from the UCI Machine Learning Repository demonstrate the effectiveness of this integration. The LLM technique excelled in producing a manageable number of if-then rules with a small number of premises while maintaining high inferential capacity for all datasets. In turn, argument-based models achieved comparable results to those derived directly from if-then rules, leveraging a concise set of rules and excelling in explainability. In summary, this paper introduces a novel approach for efficiently and automatically generating arguments and their interactions from data, addressing both scientific and applied challenges in advancing the application and deployment of argumentation systems in XAI.

We present CoCoX (short for Conceptual and Counterfactual Explanations), a model for explaining decisions made by a deep convolutional neural network (CNN). In Cognitive Psychology, the factors (or semantic-level features) that humans zoom in on when they imagine an alternative to a model prediction are often referred to as fault-lines. Motivated by this, our CoCoX model explains decisions made by a CNN using fault-lines. Specifically, given an input image I for which a CNN classification model M predicts class cpred, our fault-line based explanation identifies the minimal semantic-level features (e.g., stripes on zebra, pointed ears of dog), referred to as explainable concepts, that need to be added to or deleted from I in order to alter the classification category of I by M to another specified class calt. We argue that, due to the conceptual and counterfactual nature of fault-lines, our CoCoX explanations are practical and more natural for both expert and non-expert users to understand the internal workings of complex deep learning models. Extensive quantitative and qualitative experiments verify our hypotheses, showing that CoCoX significantly outperforms the state-of-the-art explainable AI models. Our implementation is available at https://github.com/arjunakula/CoCoX

Computational argumentation is primed to strengthen the current hot research field of Explainable Artificial Intelligence (XAI), e.g., by dialectical approaches. In this paper, we extend and discuss a recently proposed approach of so-called strong acceptance on abstract argumentation that aims to support explaining argumentative acceptance. Our goal is to push these results into the realm of structured argumentation. In this setting, a knowledge base induces an abstract argumentation framework (AF) via instantiation. We investigate how and under which conditions it is possible to transfer results regarding strong acceptance between the given knowledge base and the induced AF. To this end we consider generic functions formalizing the interaction of the AF and the knowledge base. This approach helps us to infer rather general results making basic assumptions rather than dealing with the technical details of several structured argumentation formalisms. Along the way, we apply our techniques to the concrete approach of assumption-based argumentation (ABA) which constitutes one of the primal structured argumentation formalisms.

One of the most ethically challenging issues in modern medicine are the ones related to end-of-life care. This is especially evident in intensive care setting where healthcare professionals are often...

Human cognitive and perception processes have a great tolerance for imprecision or uncertainty. For this reason, the notions of perception and cognition have great importance in solving many decision making problems in engineering, medicine, science, and social science as there are innumerable uncertainties in real-world phenomena. These uncertainties can be broadly classified as either type one uncertainty arising from the random behavior of physical processes or type two uncertainty arising from human perception and cognition processes. Statistical theory can be used to model the former, but lacks the sophistication to process the latter. The theory of fuzzy logic has proven to be very effective in processing type two uncertainty. New computing methods based on fuzzy logic can lead to greater adaptability, tractability, robustness, a lower cost solution, and better rapport with reality in the development of intelligent systems. Fuzzy logic is needed to properly pose and answer queries about quantitatively defining imprecise linguistic terms like middle class, poor, low inflation, medium inflation, and high inflation. Imprecise terms like these in natural languages should be considered to have qualitative definitions, quantitative definitions, crisp quantitative definitions, fuzzy quantitative definitions, type-one fuzzy quantitative definitions, and interval type-two fuzzy quantitative definitions. There can be crisp queries, crisp answers, type-one fuzzy queries, type-one fuzzy answers, interval type-two fuzzy queries, and interval type-two fuzzy answers.

Explainability, the aspect of artificial intelligence-based decision support (ADS) systems that allows users to understand why predictions are made, offers many potential benefits. One common claim is that explainability increases user trust, yet this has not been established in healthcare contexts. For advanced algorithms such as artificial neural networks, the generation of explanations is not trivial, but requires the use of a second algorithm. The assumption of improved user trust should therefore be investigated to determine if it justifies the additional complexity. Biochemistry staff completed a wrong blood in tube (WBIT) error identification task with the help of an ADS system. One-half of the volunteers were provided with both ADS predictions and explanations for those predictions, while the other half received predictions alone. The two groups were compared in terms of their rate of agreement with ADS predictions, as an index of user trust, and WBIT error detection performance. Since the AI model used to generate predictions was known to out-perform laboratory staff, increased trust was expected to improve user performance. Volunteers reviewed 1590 sets of results. The volunteers provided with explanations demonstrated no difference in their rate of agreement with the ADS system compared to volunteers receiving predictions alone (83.3% versus 81.8%, p = 0.46). The two volunteer groups were also equivalent in accuracy, sensitivity and specificity for WBIT error identification (p-values >0.78). For a WBIT error identification task, there was no evidence to justify the additional complexity of explainability on the grounds of increased user trust.

Statistical determination of design low flows

ChatGPT represents a transformative technology in healthcare, with demonstrated impacts across clinical practice, medical education, and research. Studies show significant efficiency gains, including 70% reduction in administrative time for discharge summaries and achievement of medical professional-level performance on standardized tests (60% accuracy on USMLE, 78.2% on PubMedQA). ChatGPT offers personalized learning platforms, automated scoring, and instant access to vast medical knowledge in medical education, addressing resource limitations and enhancing training efficiency. It streamlines clinical workflows by supporting triage processes, generating discharge summaries, and alleviating administrative burdens, allowing healthcare professionals to focus more on patient care. Additionally, ChatGPT facilitates remote monitoring and chronic disease management, providing personalized advice, medication reminders, and emotional support, thus bridging gaps between clinical visits. Its ability to process and synthesize vast amounts of data accelerates research workflows, aiding in literature reviews, hypothesis generation, and clinical trial designs. This paper aims to gather and analyze published studies involving ChatGPT, focusing on exploring its advantages and disadvantages within the healthcare context. To aid in understanding and progress, our analysis is organized into six key areas: (1) Information and Education, (2) Triage and Symptom Assessment, (3) Remote Monitoring and Support, (4) Mental Healthcare Assistance, (5) Research and Decision Support, and (6) Language Translation. Realizing ChatGPT's full potential in healthcare requires addressing key limitations, such as its lack of clinical experience, inability to process visual data, and absence of emotional intelligence. Ethical, privacy, and regulatory challenges further complicate its integration. Future improvements should focus on enhancing accuracy, developing multimodal AI models, improving empathy through sentiment analysis, and safeguarding against artificial hallucination. While not a replacement for healthcare professionals, ChatGPT can serve as a powerful assistant, augmenting their expertise to improve efficiency, accessibility, and quality of care. This collaboration ensures responsible adoption of AI in transforming healthcare delivery. While ChatGPT demonstrates significant potential in healthcare transformation, systematic evaluation of its implementation across different healthcare settings reveals varying levels of evidence quality-from robust randomized trials in medical education to preliminary observational studies in clinical practice. This heterogeneity in evidence quality necessitates a structured approach to future research and implementation.

A key ingredient of computational argumentation in AI is the generation of arguments in favor of or against claims under scrutiny. In this paper we look at the complexity of argument construction and reasoning in the prominent structured formalism of assumption-based argumentation (ABA). We point out that reasoning in ABA by means of constructing an abstract argumentation framework (AF) gives rise to two main sources of complexity: (i) constructing the AF and (ii) reasoning within the constructed graph. Since both steps are intractable in general, it is no surprise that the best performing state-of-the-art ABA reasoners skip the instantiation procedure entirely and perform tasks directly on the input knowledge base. Driven by this observation, we identify and study atomic and symmetric ABA, two ABA fragments that preserve the expressive power of general ABA, and that can be utilized to have milder complexity in the first or second step. We show that using atomic ABA allows for an instantiation procedure for general ABA leading to polynomially-bounded AFs and that symmetric ABA can be used to create AFs that have mild complexity to reason on. By an experimental evaluation, we show that using the former approach with modern AF solvers can be competitive with state-of-the-art ABA solvers, improving on previous AF instantiation approaches that are hindered by intractable argument construction.

Human decisions about the design, deployment, and governance of AI systems are increasingly complex, involving multiple desiderata such as accuracy, fairness, transparency, cost, and environmental impact. This paper proposes a novel and scalable framework for managing this complexity through application of resources from the field of Multi-Criteria Decision Analysis (MCDA). After an introduction to the “ n > 2 Criteria Problem” in decision-making about AI, an introduction to MCDA methods, and a review of prior applications of MCDA to decision support in decisions involving AI, we lay out our own novel and scalable framework for applying MCDA to such decisions. We illustrate the framework through application to four simulated scenarios involving a medical institution’s choice among AI-powered diagnostic tools with varying performance on accuracy, fairness, transparency, and cost. We apply five multi-criteria decision analysis (MCDA) methods — namely, the Weighted Sum Model (WSM), the Analytic Hierarchy Process (AHP), the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), the VIKOR method (VlseKriterijumska Optimizacija I Kompromisno Resenje), and the Preference Ranking Organization METHod for Enrichment of Evaluations (PROMETHEE) — to rank ten AI models under four distinct weight configurations reflecting realistic deployment contexts. Sensitivity analyses are then conducted to examine how variations in criterion weight assignments affect model rankings across methods and scenarios. In concluding sections we discuss the scalability of these techniques for design, deployment, and policy decisions involving AI, as well as the likely limitations of MCDA methods for decision support involving AI.

Differences of sex development (DSD) are congenital conditions in which chromosomal, gonadal, and anatomical sex characteristics are discordant with typical male or female development. These clinical practice guidelines provide evidence-based recommendations for the diagnosis and management of individuals with DSD across the lifespan. The guidelines were developed by a multidisciplinary committee of specialists representing pediatric endocrinology, adult endocrinology, urology, gynecology, psychiatry, and psychology. The committee employed a systematic review of the literature and used the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) system to assess the strength of recommendations and the quality of evidence. Key areas addressed include the initial management of infants with atypical genitalia, diagnostic approaches, hormonal treatment, surgical interventions, gonadal tumor risk assessment, fertility preservation, and the transition from pediatric to adult care. The guidelines integrate international best practices with Japan's unique sociocultural, healthcare, and legal contexts for optimal DSD management and aim to improve clinical care for individuals with DSD while acknowledging the limited high-quality evidence in many aspects of DSD management.

Graded labellings for abstract argumentation

A gradual semantics takes a weighted argumentation framework as input and outputs a final acceptability degree for each argument, with different semantics performing the computation in different manners. In this work, we consider the problem of attack inference. That is, given a gradual semantics, a set of arguments with associated initial weights, and the final desirable acceptability degrees associated with each argument, we seek to determine whether there is a set of attacks on those arguments such that we can obtain these acceptability degrees. The main contribution of our work is to demonstrate that the associated decision problem, i.e., whether a set of attacks can exist which allows the final acceptability degrees to occur for given initial weights, is NP-complete for the weighted h-categoriser and card-based semantics, and is polynomial for the weighted max-based semantics, even for the complete version of the problem (where all initial weights and final acceptability degrees are known). We then briefly discuss how this decision problem can be modified to find the attacks themselves and conclude by examining the partial problem where not all initial weights or final acceptability degrees may be known.