Interpretable Artificial Intelligence Research Articles

TinyNS: Platform-aware Neurosymbolic Auto Tiny Machine Learning

Machine learning at the extreme edge has enabled a plethora of intelligent, time-critical, and remote applications. However, deploying interpretable artificial intelligence systems that can perform high-level symbolic reasoning and satisfy the underlying system rules and physics within the tight platform resource constraints is challenging. In this article, we introduce TinyNS , the first platform-aware neurosymbolic architecture search framework for joint optimization of symbolic and neural operators. TinyNS provides recipes and parsers to automatically write microcontroller code for five types of neurosymbolic models, combining the context awareness and integrity of symbolic techniques with the robustness and performance of machine learning models. TinyNS uses a fast, gradient-free, black-box Bayesian optimizer over discontinuous, conditional, numeric, and categorical search spaces to find the best synergy of symbolic code and neural networks within the hardware resource budget. To guarantee deployability, TinyNS talks to the target hardware during the optimization process. We showcase the utility of TinyNS by deploying microcontroller-class neurosymbolic models through several case studies. In all use cases, TinyNS outperforms purely neural or purely symbolic approaches while guaranteeing execution on real hardware.

Automated abdominal CT contrast phase detection using an interpretable and open-source artificial intelligence algorithm.

To develop and validate an open-source artificial intelligence (AI) algorithm to accurately detect contrast phases in abdominal CT scans. Retrospective study aimed to develop an AI algorithm trained on 739 abdominal CT exams from 2016 to 2021, from 200 unique patients, covering 1545 axial series. We performed segmentation of five key anatomic structures-aorta, portal vein, inferior vena cava, renal parenchyma, and renal pelvis-using TotalSegmentator, a deep learning-based tool for multi-organ segmentation, and a rule-based approach to extract the renal pelvis. Radiomics features were extracted from the anatomical structures for use in a gradient-boosting classifier to identify four contrast phases: non-contrast, arterial, venous, and delayed. Internal and external validation was performed using the F1 score and other classification metrics, on the external dataset "VinDr-Multiphase CT". The training dataset consisted of 172 patients (mean age, 70 years ± 8, 22% women), and the internal test set included 28 patients (mean age, 68 years ± 8, 14% women). In internal validation, the classifier achieved an accuracy of 92.3%, with an average F1 score of 90.7%. During external validation, the algorithm maintained an accuracy of 90.1%, with an average F1 score of 82.6%. Shapley feature attribution analysis indicated that renal and vascular radiodensity values were the most important for phase classification. An open-source and interpretable AI algorithm accurately detects contrast phases in abdominal CT scans, with high accuracy and F1 scores in internal and external validation, confirming its generalization capability. Contrast phase detection in abdominal CT scans is a critical step for downstream AI applications, deploying algorithms in the clinical setting, and for quantifying imaging biomarkers, ultimately allowing for better diagnostics and increased access to diagnostic imaging. Digital Imaging and Communications in Medicine labels are inaccurate for determining the abdominal CT scan phase. AI provides great help in accurately discriminating the contrast phase. Accurate contrast phase determination aids downstream AI applications and biomarker quantification.

Crafting explainable artificial intelligence through active inference: A model for transparent introspection and decision-making

This paper explores the feasibility of constructing interpretable artificial intelligence (AI) systems rooted in active inference and the free energy principle. Initially, we offer a concise introduction to active inference, emphasizing its relevance to modeling decision-making, introspection, and the generation of both overt and covert actions. Subsequently, we delve into how active inference can serve as a foundation for designing explainable AI systems. Specifically, it enables us to capture essential aspects of "introspective" processes and generate intelligible models of decision-making mechanisms. We propose an architectural framework for explainable AI systems employing active inference. Central to this framework is an explicit hierarchical generative model that enables the AI system to monitor and elucidate the factors influencing its decisions. Importantly, this model's structure is designed to be understandable and verifiable by human users. We elucidate how this architecture can amalgamate diverse data sources to make informed decisions in a transparent manner, mirroring aspects of human consciousness and introspection. Finally, we examine the implications of our findings for future AI research and discuss potential ethical considerations associated with developing AI systems with (apparent) introspective capabilities.

Crafting explainable artificial intelligence through active inference: A model for transparent introspection and decision-making

This paper explores the feasibility of constructing interpretable artificial intelligence (AI) systems rooted in active inference and the free energy principle. Initially, we offer a concise introduction to active inference, emphasizing its relevance to modeling decision-making, introspection, and the generation of both overt and covert actions. Subsequently, we delve into how active inference can serve as a foundation for designing explainable AI systems. Specifically, it enables us to capture essential aspects of "introspective" processes and generate intelligible models of decision-making mechanisms. We propose an architectural framework for explainable AI systems employing active inference. Central to this framework is an explicit hierarchical generative model that enables the AI system to monitor and elucidate the factors influencing its decisions. Importantly, this model's structure is designed to be understandable and verifiable by human users. We elucidate how this architecture can amalgamate diverse data sources to make informed decisions in a transparent manner, mirroring aspects of human consciousness and introspection. Finally, we examine the implications of our findings for future AI research and discuss potential ethical considerations associated with developing AI systems with (apparent) introspective capabilities.

A novel framework for high resolution air quality index prediction with interpretable artificial intelligence and uncertainties estimation

Accurate air quality index (AQI) prediction is essential in environmental monitoring and management. Given that previous studies neglect the importance of uncertainty estimation and the necessity of constraining the output during prediction, we proposed a new hybrid model, namely TMSSICX, to forecast the AQI of multiple cities. Firstly, time-varying filtered based empirical mode decomposition (TVFEMD) was adopted to decompose the AQI sequence into multiple internal mode functions (IMF) components. Secondly, multi-scale fuzzy entropy (MFE) was applied to evaluate the complexity of each IMF component and clustered them into high and low-frequency portions. In addition, the high-frequency portion was secondarily decomposed by successive variational mode decomposition (SVMD) to reduce volatility. Then, six air pollutant concentrations, namely CO, SO2, PM2.5, PM10, O3, and NO2, were used as inputs. The secondary decomposition and preliminary portion were employed as the outputs for the bidirectional long short-term memory network optimized by the snake optimization algorithm (SOABiLSTM) and improved Catboost (ICatboost), respectively. Furthermore, extreme gradient boosting (XGBoost) was applied to ensemble each predicted sub-model to acquire the consequence. Ultimately, we introduced adaptive kernel density estimation (AKDE) for interval estimation. The empirical outcome indicated the TMSSICX model achieved the best performance among the other 23 models across all datasets. Moreover, implementing the XGBoost to ensemble each predicted sub-model led to an 8.73%, 8.94%, and 0.19% reduction in RMSE, compared to SVM. Additionally, by utilizing SHapley Additive exPlanations (SHAP) to assess the impact of the six pollutant concentrations on AQI, the results reveal that PM2.5 and PM10 had the most notable positive effects on the long-term trend of AQI. We hope this model can provide guidance for air quality management.

Application of Artificial Intelligence in Healthcare: The Need for More Interpretable Artificial Intelligence.

N/a.

Unlocking Machine Learning Model Decisions: A Comparative Analysis of LIME and SHAP for Enhanced Interpretability

XAI is critical for establishing trust and enabling the appropriate development of machine learning models. By offering transparency into how these models make judgements, XAI enables researchers and users to uncover potential biases, admit limits, and eventually enhance the fairness and dependability of AI systems. In this paper, we demonstrates two techniques, LIME and SHAP, used to improve the interpretability of machine learning models. Assessing Explainable AI (XAI) approaches is critical in searching for transparent and interpretable artificial intelligence (AI) models. Explainable AI (XAI) approaches are designed to provide insight into how complex models make decisions. This paper thoroughly analyzes two prominent XAI methods: Shapley Additive explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME). This study aims to understand the decision made by a machine learning model and how the model came to that decision. We discuss the approaches and framework of both LIME and SHAP and assess their behavior in predicting the model's outcome.

Explainable artificial intelligence models for predicting risk of suicide using health administrative data in Quebec.

Suicide is a complex, multidimensional event, and a significant challenge for prevention globally. Artificial intelligence (AI) and machine learning (ML) have emerged to harness large-scale datasets to enhance risk detection. In order to trust and act upon the predictions made with ML, more intuitive user interfaces must be validated. Thus, Interpretable AI is one of the crucial directions which could allow policy and decision makers to make reasonable and data-driven decisions that can ultimately lead to better mental health services planning and suicide prevention. This research aimed to develop sex-specific ML models for predicting the population risk of suicide and to interpret the models. Data were from the Quebec Integrated Chronic Disease Surveillance System (QICDSS), covering up to 98% of the population in the province of Quebec and containing data for over 20,000 suicides between 2002 and 2019. We employed a case-control study design. Individuals were considered cases if they were aged 15+ and had died from suicide between January 1st, 2002, and December 31st, 2019 (n = 18339). Controls were a random sample of 1% of the Quebec population aged 15+ of each year, who were alive on December 31st of each year, from 2002 to 2019 (n = 1,307,370). We included 103 features, including individual, programmatic, systemic, and community factors, measured up to five years prior to the suicide events. We trained and then validated the sex-specific predictive risk model using supervised ML algorithms, including Logistic Regression (LR), Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Multilayer perceptron (MLP). We computed operating characteristics, including sensitivity, specificity, and Positive Predictive Value (PPV). We then generated receiver operating characteristic (ROC) curves to predict suicides and calibration measures. For interpretability, Shapley Additive Explanations (SHAP) was used with the global explanation to determine how much the input features contribute to the models' output and the largest absolute coefficients. The best sensitivity was 0.38 with logistic regression for males and 0.47 with MLP for females; the XGBoost Classifier with 0.25 for males and 0.19 for females had the best precision (PPV). This study demonstrated the useful potential of explainable AI models as tools for decision-making and population-level suicide prevention actions. The ML models included individual, programmatic, systemic, and community levels variables available routinely to decision makers and planners in a public managed care system. Caution shall be exercised in the interpretation of variables associated in a predictive model since they are not causal, and other designs are required to establish the value of individual treatments. The next steps are to produce an intuitive user interface for decision makers, planners and other stakeholders like clinicians or representatives of families and people with live experience of suicidal behaviors or death by suicide. For example, how variations in the quality of local area primary care programs for depression or substance use disorders or increased in regional mental health and addiction budgets would lower suicide rates.

Guest Editorial: New Developments in Explainable and Interpretable Artificial Intelligence

Guest Editorial: New Developments in Explainable and Interpretable Artificial Intelligence

Unlocking Machine Learning Model Decisions: A Comparative Analysis of LIME and SHAP for Enhanced Interpretability

XAI is critical for establishing trust and enabling the appropriate development of machine learning models. By offering transparency into how these models make judgements, XAI enables researchers and users to uncover potential biases, admit limits, and eventually enhance the fairness and dependability of AI systems. In this paper, we demonstrates two techniques, LIME and SHAP, used to improve the interpretability of machine learning models. Assessing Explainable AI (XAI) approaches is critical in searching for transparent and interpretable artificial intelligence (AI) models. Explainable AI (XAI) approaches are designed to provide insight into how complex models make decisions. This paper thoroughly analyzes two prominent XAI methods: Shapley Additive explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME). This study aims to understand the decision made by a machine learning model and how the model came to that decision. We discuss the approaches and framework of both LIME and SHAP and assess their behavior in predicting the model's outcome.

Clinical Features Predicting COVID-19 Severity Risk at the Time of Hospitalization.

The global spread of COVID-19 has led to significant mortality and morbidity worldwide. Early identification of COVID-19 patients who are at high risk of developing severe disease can help in improved patient management, care, and treatment, as well as in the effective allocation of hospital resources. The severity prediction at the time of hospitalization can be extremely helpful in deciding the treatment of COVID-19 patients. To this end, this study presents an interpretable artificial intelligence (AI) model, named COVID-19 severity predictor (CoSP) that predicts COVID-19 severity using the clinical features at the time of hospital admission. We utilized a dataset comprising 64 demographic and laboratory features of 7,416 confirmed COVID-19 patients that were collected at the time of hospital admission. The proposed hierarchical CoSP model performs four-class COVID severity risk prediction into asymptomatic, mild, moderate, and severe categories. CoSP yielded better performance with good interpretability, as observed via Shapley analysis on COVID severity prediction compared to the other popular ML methods, with an area under the received operating characteristic curve (AUC-ROC) of 0.95, an area under the precision-recall curve (AUPRC) of 0.91, and a weighted F1-score of 0.83. Out of 64 initial features, 19 features were inferred as predictive of the severity of COVID-19 disease by the CoSP model. Therefore, an AI model predicting COVID-19 severity may be helpful for early intervention, optimizing resource allocation, and guiding personalized treatments, potentially enabling healthcare professionals to save lives and allocate resources effectively in the fight against the pandemic.

Artificial intelligence in analogue circuit design

Analog circuit design plays an integral part in modern society; however, traditional approaches often fall short in terms of efficiency and effectiveness. This paper explores the application and challenges presented by artificial intelligence (AI) technologies in analog circuit design. The objective of this paper was to evaluate the capacity of artificial intelligence technologies such as machine learning, deep learning and reinforcement learning to advance circuit modeling, optimization and layout design processes. A comprehensive literature review and analysis were performed. The study finds that while AI can make significant contributions to circuit design, several challenges must first be overcome. These include data scarcity for training AI models and the opaque nature of AI decisions which prevent their widespread acceptance. Finally, future research directions such as creating synthetic data or developing more interpretable AI models were identified as potential future research avenues.

An Interpretable Time Series Forecasting Model for Predicting NOx Emission Concentration in Ferroalloy Electric Arc Furnace Plants

Considering the pivotal role of ferroalloys in the steel industry and the escalating global emphasis on sustainability (e.g., zero emissions and carbon neutrality), the demand for ferroalloys is anticipated to increase. However, the electric arc furnace (EAF) of ferroalloy plants generates substantial amounts of nitrogen oxides (NOx) because of the high-temperature combustion processes. Despite the substantial contributions of many studies on NOx prediction from various industrial facilities, there is a lack of studies considering the environmental condition of the EAF in ferroalloy plants. Therefore, this study presents a deep learning model for predicting NOx emissions from ferroalloy plants and further can provide guidelines for predicting NOx in industrial sites equipped with electric furnaces. In this study, we collected various historical data from the manufacturing execution system of electric furnaces and exhaust gas systems to develop a prediction model. Additionally, an interpretable artificial intelligence method was employed to track the effects of each variable on the NOx emissions. The proposed prediction model can provide decision support to reduce NOx emissions. Furthermore, the interpretation of the model contributes to a better understanding of the factors influencing NOx emissions and the development of effective strategies for emission reduction in ferroalloys EAF plants.

Augmenting Deep Neural Networks with Symbolic Educational Knowledge: Towards Trustworthy and Interpretable AI for Education

Artificial neural networks (ANNs) have proven to be among the most important artificial intelligence (AI) techniques in educational applications, providing adaptive educational services. However, their educational potential is limited in practice due to challenges such as the following: (i) the difficulties in incorporating symbolic educational knowledge (e.g., causal relationships and practitioners’ knowledge) in their development, (ii) a propensity to learn and reflect biases, and (iii) a lack of interpretability. As education is classified as a ‘high-risk’ domain under recent regulatory frameworks like the EU AI Act—highlighting its influence on individual futures and discrimination risks—integrating educational insights into ANNs is essential. This ensures that AI applications adhere to essential educational restrictions and provide interpretable predictions. This research introduces NSAI, a neural-symbolic AI approach that integrates neural networks with knowledge representation and symbolic reasoning. It injects and extracts educational knowledge into and from deep neural networks to model learners’ computational thinking, aiming to enhance personalized learning and develop computational thinking skills. Our findings revealed that the NSAI approach demonstrates better generalizability compared to deep neural networks trained on both original training data and data enriched by SMOTE and autoencoder methods. More importantly, we found that, unlike traditional deep neural networks, which mainly relied on spurious correlations in their predictions, the NSAI approach prioritizes the development of robust representations that accurately capture causal relationships between inputs and outputs. This focus significantly reduces the reinforcement of biases and prevents misleading correlations in the models. Furthermore, our research showed that the NSAI approach enables the extraction of rules from the trained network, facilitating interpretation and reasoning during the path to predictions, as well as refining the initial educational knowledge. These findings imply that neural-symbolic AI not only overcomes the limitations of ANNs in education but also holds broader potential for transforming educational practices and outcomes through trustworthy and interpretable applications.

Technical and functional design considerations for a real-world interpretable AI solution for NIR perfusion analysis (including cancer)

Near infrared (NIR) analysis of tissue perfusion via indocyanine green fluorescence assessment is performed clinically during surgery for a range of indications. Its usefulness can potentially be further enhanced through the application of interpretable artificial intelligence (AI) methods to improve dynamic interpretation accuracy in these and also open new applications. While its main use currently is for perfusion assessment as a tissue health check prior to performing an anastomosis, there is increasing interest in using fluorophores for cancer detection during surgical interventions with most research being based on the paradigm of static imaging for fluorophore uptake hours after preoperative dosing. Although some image boosting and relative estimation of fluorescence signals is already inbuilt into commercial NIR systems, fuller implementation of AI methods can enable actionable predictions especially when applied during the dynamic, early inflow-outflow phase that occurs seconds to minutes after ICG (or indeed other fluorophore) administration. Already research has shown that such methods can accurately differentiate cancer from benign tissue in the operating theatre in real time in principle based on their differential signalling and could be useful for tissue perfusion classification more generally. This can be achieved through the generation of fluorescence intensity curves from an intra-operative NIR video stream. These curves are processed to adjust for image disturbances and curve features known to be influential in tissue characterisation are extracted. Existing machine learning based classifiers can then use these features to classify the tissue in question according to prior training sets. The use of this interpretable methodology enables accurate classification algorithms to be built with modest training sets in comparison to those required for deep learning modelling in addition to achieving compliance with medical device regulations. Integration of the multiple algorithms required to achieve this classification into a desktop application or medical device could make the use of this method accessible and useful to (as well as useable by) surgeons without prior training in computer technology. This document details some technical and functional design considerations underlying such a novel recommender system to advance the foundational concept and methodology as software as medical device for in situ cancer characterisation with relevance more broadly also to other tissue perfusion applications.

Enhancing Transparency and Interpretability in Deep Learning Models: A Comprehensive Study on Explainable AI Techniques

Abstract: Deep learning models have demonstrated remarkable capabilities across various domains, but their inherent complexity often leads to challenges in understanding and interpreting their decisions. The demand for transparent and interpretable artificial intelligence (AI) systems is particularly crucial in fields such as healthcare, finance, and autonomous systems. This research paper presents a comprehensive study on the application of Explainable AI (XAI) techniques to enhance transparency and interpretability in deep learning models. Keywords: Explainable AI (XAI), artificial intelligence (AI).

Enhancing Trust in Healthcare: The Role of AI Explainability and Professional Familiarity

The integration of Artificial Intelligence (AI) in healthcare has been impeded by a significant issue: a lack of trust among healthcare professionals, stemming from the opacity of AI decision-making processes and a general unfamiliarity with AI technologies. This study investigates the impact of AI's explainability and healthcare professionals' familiarity with AI on their trust in AI applications within healthcare settings. Adopting a quantitative research methodology, the study utilized a structured questionnaire to gather data from a diverse group of healthcare professionals, including doctors, nurses, and administrators, across various hospitals and healthcare institutions in Pakistan. The research employed a stratified random sampling approach to ensure a comprehensive and representative data set. The results indicated a positive and significant relationship between AI explainability and trust in AI (Path Coefficient: 0.62, t-Value: 5.20), suggesting that clearer and more transparent AI decision-making processes enhance healthcare professionals' trust., Similarly, familiarity with AI was found to positively influence trust in AI (Path Coefficient: 0.48, t-Value: 4.35), highlighting the importance of exposure and understanding of AI systems among healthcare professionals. These findings have crucial implications for both AI developers and healthcare administrators. For AI developers, the emphasis must be on creating more transparent and interpretable AI systems.  For healthcare administrators, the results suggest the need to invest in training and educational programs to increase professionals' familiarity with AI, thereby enhancing trust and acceptance.  The study significantly contributes to the existing literature by empirically validating the importance of AI explainability and familiarity in building trust in AI within the healthcare context, especially in a developing country setting. For policymakers, these insights are critical in guiding strategies and policies aimed at effectively integrating AI into healthcare systems. By addressing the identified factors, healthcare sectors can better leverage AI's potential, leading to improved patient care and more efficient healthcare operations.

Artificial intelligence (AI)-it's the end of the tox as we know it (and I feel fine).

The rapid progress of AI impacts diverse scientific disciplines, including toxicology, and has the potential to transform chemical safety evaluation. Toxicology has evolved from an empirical science focused on observing apical outcomes of chemical exposure, to a data-rich field ripe for AI integration. The volume, variety and velocity of toxicological data from legacy studies, literature, high-throughput assays, sensor technologies and omics approaches create opportunities but also complexities that AI can help address. In particular, machine learning is well suited to handle and integrate large, heterogeneous datasets that are both structured and unstructured-a key challenge in modern toxicology. AI methods like deep neural networks, large language models, and natural language processing have successfully predicted toxicity endpoints, analyzed high-throughput data, extracted facts from literature, and generated synthetic data. Beyond automating data capture, analysis, and prediction, AI techniques show promise for accelerating quantitative risk assessment by providing probabilistic outputs to capture uncertainties. AI also enables explanation methods to unravel mechanisms and increase trust in modeled predictions. However, issues like model interpretability, data biases, and transparency currently limit regulatory endorsement of AI. Multidisciplinary collaboration is needed to ensure development of interpretable, robust, and human-centered AI systems. Rather than just automating human tasks at scale, transformative AI can catalyze innovation in how evidence is gathered, data are generated, hypotheses are formed and tested, and tasks are performed to usher new paradigms in chemical safety assessment. Used judiciously, AI has immense potential to advance toxicology into a more predictive, mechanism-based, and evidence-integrated scientific discipline to better safeguard human and environmental wellbeing across diverse populations.

Special issue on “Towards robust explainable and interpretable artificial intelligence”

Special issue on “Towards robust explainable and interpretable artificial intelligence”

Interpretable Artificial Intelligence in Information Systems: Status Review and Future Research Directions

Efforts to develop black-box artificial intelligence (AI) systems have become a phenomenon of emerging global interest in academia, business, and society, and have led to the development of the XAI research field. With its pluralistic perspective, information systems (IS) research is destined to contribute to this emerging field; thus, it is not surprising that the number of research publications at XAI has increased significantly. This paper aims to provide a comprehensive overview of XAI research in public and electronic markets, specifically using a structured literature review. Based on a literature review of 180 research papers, this work examines the most receptive points, the development of academic debates, and the most important concepts and methodologies. In addition, eight research areas with different levels of maturity in e-markets are identified. Finally, guidelines for the XAI research agenda in IS are presented