1,720,987 research outputs found
HUMAN-CENTRED COUNTERFACTUAL EXPLANATIONS FOR EXPLAINABLE ARTIFICIAL INTELLIGENCE
No full textExplainable AI (XAI) research endeavors to enhance transparency and interpretability in Artificial Intelligence (AI) systems by elucidating their decision-making processes. However, existing XAI methods often fall short in meeting user requirements, leading to challenges such as overly complex explanations and mismatches between user expectations and provided explanations. Human-centred XAI (HC-XAI) emerges as a solution to these challenges, focusing on designing explanations that are actionable, user-friendly, and customizable to individual preferences. This dissertation proposes a human-centred approach to XAI, aiming to bridge the gap between technical advancements and practical usability. The research involves conceptualizing and implementing an XAI framework that provides transparent and understandable explanations to users. Existing algorithms often operate within the entire feature space when looking for counterfactuals (i.e., alternative facts to the observed ones) with the purpose of optimizing changes to address undesired outcomes, overlooking the identification of key contributors and practicality of suggested changes. To overcome these limitations and enhance the confidence in provided explanations, this dissertation presents a new approach for generating user feedback-based counterfactual explanations (UFCE). UFCE allows for the incorporation of user constraints to determine minimal modifications in actionable feature subsets while considering feature dependence. UFCE is a model-agnostic method for tabular datasets. UFCE is crafted by looking at human needs and caters to such needs by giving the opportunity to end users to customize the explanations for machine learning predictions. UFCE is developed as open-source software to promote open science. To evaluate the effectiveness of the developed approach, a novel interactive web-based game is implemented, allowing users to engage with the XAI system and receive explanations for machine learning decision-making tasks. Task performance, usability, and user satisfaction assessments are conducted to provide comprehensive insights into the practical applicability of UFCE, the new HC-XAI approach in real-world decision-making scenarios. Through this research, we aim to foster greater acceptance and utilization of AI technologies across various domains.Explainable AI (XAI) research endeavors to enhance transparency and interpretability in Artificial Intelligence (AI) systems by elucidating their decision-making processes. However, existing XAI methods often fall short in meeting user requirements, leading to challenges such as overly complex explanations and mismatches between user expectations and provided explanations. Human-centred XAI (HC-XAI) emerges as a solution to these challenges, focusing on designing explanations that are actionable, user-friendly, and customizable to individual preferences. This dissertation proposes a human-centred approach to XAI, aiming to bridge the gap between technical advancements and practical usability. The research involves conceptualizing and implementing an XAI framework that provides transparent and understandable explanations to users. Existing algorithms often operate within the entire feature space when looking for counterfactuals (i.e., alternative facts to the observed ones) with the purpose of optimizing changes to address undesired outcomes, overlooking the identification of key contributors and practicality of suggested changes. To overcome these limitations and enhance the confidence in provided explanations, this dissertation presents a new approach for generating user feedback-based counterfactual explanations (UFCE). UFCE allows for the incorporation of user constraints to determine minimal modifications in actionable feature subsets while considering feature dependence. UFCE is a model-agnostic method for tabular datasets. UFCE is crafted by looking at human needs and caters to such needs by giving the opportunity to end users to customize the explanations for machine learning predictions. UFCE is developed as open-source software to promote open science. To evaluate the effectiveness of the developed approach, a novel interactive web-based game is implemented, allowing users to engage with the XAI system and receive explanations for machine learning decision-making tasks. Task performance, usability, and user satisfaction assessments are conducted to provide comprehensive insights into the practical applicability of UFCE, the new HC-XAI approach in real-world decision-making scenarios. Through this research, we aim to foster greater acceptance and utilization of AI technologies across various domains
Towards Human Cognition Level-based Experiment Design for Counterfactual Explanations (XAI)
Full text linkExplainable Artificial Intelligence (XAI) has recently gained a swell of
interest, as many Artificial Intelligence (AI) practitioners and developers are
compelled to rationalize how such AI-based systems work. Decades back, most XAI
systems were developed as knowledge-based or expert systems. These systems
assumed reasoning for the technical description of an explanation, with little
regard for the user's cognitive capabilities. The emphasis of XAI research
appears to have turned to a more pragmatic explanation approach for better
understanding. An extensive area where cognitive science research may
substantially influence XAI advancements is evaluating user knowledge and
feedback, which are essential for XAI system evaluation. To this end, we
propose a framework to experiment with generating and evaluating the
explanations on the grounds of different cognitive levels of understanding. In
this regard, we adopt Bloom's taxonomy, a widely accepted model for assessing
the user's cognitive capability. We utilize the counterfactual explanations as
an explanation-providing medium encompassed with user feedback to validate the
levels of understanding about the explanation at each cognitive level and
improvise the explanation generation methods accordingly.Comment: 5 pages, 2 figure
How to Build Self-Explaining Fuzzy Systems: From Interpretability to Explainability [AI-eXplained]
No full textFCE: Feedback Based Counterfactual Explanations for Explainable AI
Full text linkArtificial Intelligence can provide quite accurate predictions for critical applications
(e.g., healthcare), but lacks the ability to explain its internal mechanism in most applications which
require high interaction with humans. Even if many studies analyze machine learning models and their
learning behavior and eventually provide an interpretation of the inner mechanics of these models, these
studies often entail a simpler surrogate model, which generates explanations by producing a piece of
interpretable information such as feature scores. The crucial caveat against these studies is the lack of
human involvement in the design and evaluation of explanations, consequently giving rise to trust issues
and lack of acceptance and understanding. To this end, we address this limitation by involving humans in
the counterfactual explanation generation process which is enriched with user feedback, thus enhancing the
automated explanations which are better aligned with user expectations. In this paper, we propose a user
feedback based counterfactual explanation approach (FCE) for explainable Artificial Intelligence. In our
work, we utilize feedback in two ways: first, to customize the explanations by providing the acceptable
ranges in the feature space where to look for feasible counterfactuals, and second, to evaluate the generated
explanations
Introducing User Feedback-Based Counterfactual Explanations (UFCE)
No full textAbstract Machine learning models are widely used in real-world applications. However, their complexity makes it often challenging to interpret the rationale behind their decisions. Counterfactual explanations (CEs) have emerged as a viable solution for generating comprehensible explanations in eXplainable Artificial Intelligence (XAI). CE provides actionable information to users on how to achieve the desired outcome with minimal modifications to the input. However, current CE algorithms usually operate within the entire feature space when optimising changes to turn over an undesired outcome, overlooking the identification of key contributors to the outcome and disregarding the practicality of the suggested changes. In this study, we introduce a novel methodology, that is named as user feedback-based counterfactual explanation (UFCE), which addresses these limitations and aims to bolster confidence in the provided explanations. UFCE allows for the inclusion of user constraints to determine the smallest modifications in the subset of actionable features while considering feature dependence, and evaluates the practicality of suggested changes using benchmark evaluation metrics. We conducted three experiments with five datasets, demonstrating that UFCE outperforms two well-known CE methods in terms of proximity, sparsity, and feasibility. Reported results indicate that user constraints influence the generation of feasible CEs
Gender Classification Using Smartphone Sensors and Machine Learning Approaches
No full textGait analysis is typically associated with the pattern of the human walk. Determining it with computational means can be helpful in many ways-from identifying individual humans to detecting gait-related diseases. In comparison to the expensive approaches and devices, which are limited to laboratories, smart- phones with motion sensors are low-cost solutions through which we can analyze mobility and gait patterns. Thus, in this work, we present the usage of smartphone sensors for data acquisition followed by machine learning-based gender classification, which is a baseline for different gait-related tasks. In this regard, we collected data from 14 persons in different tracks, paces, and movement styles; after adequate normalization, iterative feature elimination, and Monte-Carlo experiment-based ML training, we found the Decision Tree is the most optimal algorithm with attaining 90.6 % balanced-accuracy
Predicting metabolic responses in genetic disorders via structural representation in machine learning
No full textMetabolomics has emerged as a promising discipline in pharmaceuticals and preventive healthcare. However, analysing large metabolomics datasets remains challenging due to limited and incompletely annotated biological pathways. To address this limitation, we recently proposed training machine learning classifiers on molecular fingerprints of metabolites to predict their responses under specific conditions and analysing feature importance to identify key chemical configurations, providing insights into the affected biological processes. This study extends our previous research by evaluating various metabolite structural representations, including Morgan fingerprint and its variants, graph-based structural encodings and proposing novel representations to improve resolution and interpretability of the state-of-the-art approaches. These structural encodings were evaluated on mass spectrometry metabolomic data for a cellular model of the genetic disease Ataxia Telangiectasia. The study found that machine learning classifiers trained on the new representations improved in classification accuracy and interpretability. Notably, models trained on graph-based encoding do not exhibit performance gains, not even with pre-training on a larger metabolite dataset, underlining the efficacy of our proposed representations. Finally, feature importance analysis across different encoding methods consistently identifies similar structures as relevant for classification, underscoring the robustness of our approach across diverse structural representations
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Interpretable Machine Learning for Automated Cellular Population Analysis in Flow Cytometry
No full textGiven the significant improvement in flow cytometry technologies, a massive dataset has been acquired, measuring a larger number of cellular markers. Managing and analysing these data by hand, based on the sole expertise of human operators, is no longer feasible. Recent literature suggests exploiting machine learning algorithms to automate data analysis for extracting and counting sub-populations of cells and supporting diagnosis. In this paper, we applied Support Vector Machine, XGBoost, Decision Tree, Logistic Regression, and Multi-Layer Perceptron to identify three specific cellular types: Lymphocyte T, B, and T cytotoxic. Performances are promising across all models and experiments, with a balanced accuracy above 0.85. Moreover, when looking at the Recall and the F1 score, the Decision Tree is the unique model with values below 0.8 in the classification B Lymphocyte. Moreover, to improve the interpretability of the trained models, we computed the SHAP-based explanations for the XGBoost, Decision Tree and Multi-Layer Perceptron, obtaining a set of extracted features that domain experts recognised as significant for the three classification tasks, thus emphasising the viability of this approach in automating the gating process in flow citometry
- …
