Bias Mitigation Research Articles

The Adaptive Human-AI Synergy in Logistics (AHASL) Theory

This paper introduces the Adaptive Human-AI Synergy in Logistics (AHASL) theory, which focuses on integrating artificial intelligence (AI) into logistics and supply chain management. The study takes a qualitative approach, using interviews, observations, and reflexive journaling, and finds that while AI significantly improves operational efficiency and decision-making, human oversight is still required to address AI's limitations, such as bias, lack of transparency, and potential skill erosion. The AHASL model prioritizes Full-Spectrum Explainability (FSE) and a Bias Mitigation Framework (BMF) to ensure that AI-driven decisions are transparent and equitable while simultaneously arguing for retaining human expertise to retain adaptability and resilience in logistical operations. The study's findings emphasize the significance of balanced human-AI collaboration. However, they also call for more research into the long-term effects, scalability, and ethical implications of AI integration in logistics.

Sex differences in brain MRI using deep learning toward fairer healthcare outcomes

This study leverages deep learning to analyze sex differences in brain MRI data, aiming to further advance fairness in medical imaging. We employed 3D T1-weighted Magnetic Resonance images from four diverse datasets: Calgary-Campinas-359, OASIS-3, Alzheimer's Disease Neuroimaging Initiative, and Cambridge Center for Aging and Neuroscience, ensuring a balanced representation of sexes and a broad demographic scope. Our methodology focused on minimal preprocessing to preserve the integrity of brain structures, utilizing a Convolutional Neural Network model for sex classification. The model achieved an accuracy of 87% on the test set without employing total intracranial volume (TIV) adjustment techniques. We observed that while the model exhibited biases at extreme brain sizes, it performed with less bias when the TIV distributions overlapped more. Saliency maps were used to identify brain regions significant in sex differentiation, revealing that certain supratentorial and infratentorial regions were important for predictions. Furthermore, our interdisciplinary team, comprising machine learning specialists and a radiologist, ensured diverse perspectives in validating the results. The detailed investigation of sex differences in brain MRI in this study, highlighted by the sex differences map, offers valuable insights into sex-specific aspects of medical imaging and could aid in developing sex-based bias mitigation strategies, contributing to the future development of fair AI algorithms. Awareness of the brain's differences between sexes enables more equitable AI predictions, promoting fairness in healthcare outcomes. Our code and saliency maps are available at https://github.com/mahsadibaji/sex-differences-brain-dl.

Indirect Treatment Comparisons in Healthcare Decision Making: A Targeted Review of Regulatory Approval, Reimbursement, and Pricing Recommendations Globally for Oncology Drugs in 2021-2023.

Indirect treatment comparisons (ITCs) evaluate novel treatments compared to appropriate comparators when direct evidence is unavailable or infeasible. The objective of this study was to highlight the prevalence of different ITC methods in oncology drug submissions and to provide insights into how ITCs have been used in recent regulatory approval, reimbursement recommendations, or pricing decisions across various regions and diverse assessment frameworks. A targeted literature review was conducted to identify assessment documents for oncology drug submissions that included ITCs. This included hand searches of the websites of four regulatory bodies and four health technology assessment (HTA) agencies with varying assessment frameworks across North America, Europe, and Asia-Pacific. A total of 185 documents were included for synthesis. Documents were retrieved from all four HTA agencies and the European Medicines Agency (EMA), the only regulatory body with eligible records. Within these, 188 unique submissions included a total of 306 supporting ITCs of various methods. Authorities more frequently favored anchored or population-adjusted ITC techniques for their effectiveness in data adjustment and bias mitigation. Furthermore, ITCs in orphan drug submissions more frequently led to positive decisions compared to non-orphan submissions. This review highlights the crucial role and widespread use of ITCs in global healthcare decision-making, particularly when direct evidence is lacking, and in the discernment of market-specific clinical benefits. This work contributes to bolstering the credibility and recognition of ITCs across regulatory and HTA agencies of diverse regions and assessment frameworks.

Empowering big data analytics through machine learning: Applications, challenges, and future directions

Abstract. This paper explores the transformative role of machine learning (ML) methodologies in big data analytics, highlighting supervised learning, unsupervised learning, and deep neural networks' contributions to diverse sectors. Supervised learning, with regression analysis at its core, provides accurate forecasting in finance and healthcare by modeling relationships between variables. Unsupervised learning, through techniques like k-means and hierarchical clustering, uncovers patterns within data, offering insights for retail and biological analysis. Deep learning, particularly Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) with Long Short-Term Memory (LSTM), excel in complex tasks like image recognition and sequential data processing, driving advances in fields from autonomous driving to language translation. The application of ML in enhancing business intelligence, innovating fintech, and advancing healthcare analytics is discussed, alongside the challenges of data privacy, security, ethical considerations, and the skills gap. This paper underscores the need for advanced cryptographic techniques, bias mitigation strategies, and education to address these challenges. It concludes by emphasizing the importance of interdisciplinary education and AI advancements in bridging the skills gap, ensuring the ethical use of ML, and making these technologies accessible for future innovations.

Bias and Its Consequences : A Study of Machine Learning Performance

This paper addresses the concern about bias affecting the results of machine learning models. For this purpose, it uses the Adult Income dataset from OpenML for income classification. The conditions for bias are induced by underrepresenting people that earn <= $50K in training data, thus checking the behavior of different models when encountering such a skewed distribution. Key metrics, namely accuracy and specificity (True Negative Rate), were analyzed for unbiased and biased training scenarios. The results show that Naive Bayes and Random Forest models were resistant to bias, but others, including SVM and Logistic Regression, suffered major performance drops. This study throws light on the robustness of different classifiers when exposed to biased data, requiring further bias mitigation strategies in real-world applications. This paper actually examines critically how bias in training data can significantly affect the performance of prediction, fairness, and model selection in income classification tasks.

The Self-Improving Customer Service System: How AI Learns from Every Interaction

This article explores the transformative impact of self-improving AI systems in customer service, examining the parallels between AI and human learning processes, the mechanics of AI self-improvement, and the advantages of AI's learning at scale. It delves into real-world applications and impacts across various industries, highlighting improvements in personalization, efficiency, accuracy, customer satisfaction, and cost reduction. The article also addresses ethical considerations such as data privacy, transparency, and bias mitigation, while offering a future outlook on advanced AI capabilities in emotional intelligence, predictive service, and seamless human-AI collaboration. Through a comprehensive analysis of current technologies and future possibilities, this study provides insights into how self-improving AI is revolutionizing customer service, promising more efficient, personalized, and satisfying experiences for customers.

Bias Sensitivity in Diagnostic Decision-Making: Comparing ChatGPT with Residents.

Diagnostic errors, often due to biases in clinical reasoning, significantly affect patient care. While artificial intelligence chatbots like ChatGPT could help mitigate such biases, their potential susceptibility to biases is unknown. This study evaluated diagnostic accuracy of ChatGPT against the performance of 265 medical residents in five previously published experiments aimed at inducing bias. The residents worked in several major teaching hospitals in the Netherlands. The biases studied were case-intrinsic (presence of salient distracting findings in the patient history, effects of disruptive patient behaviors) and situational (prior availability of a look-alike patient). ChatGPT's accuracy in identifying the most-likely diagnosis was measured. Diagnostic accuracy of residents and ChatGPT was equivalent. For clinical cases involving case-intrinsic bias, both ChatGPT and the residents exhibited a decline in diagnostic accuracy. Residents' accuracy decreased on average 12%, while the accuracy of ChatGPT 4.0 decreased 21%. Accuracy of ChatGPT 3.5 decreased 9%. These findings suggest that, like human diagnosticians, ChatGPT is sensitive to bias when the biasing information is part of the patient history. When the biasing information was extrinsic to the case in the form of the prior availability of a look-alike case, residents' accuracy decreased by 15%. By contrast, ChatGPT's performance was not affected by the biasing information. Chi-square goodness-of-fit tests corroborated these outcomes. It seems that, while ChatGPT is not sensitive to bias when biasing information is situational, it is sensitive to bias when the biasing information is part of the patient's disease history. Its utility in diagnostic support has potential, but caution is advised. Future research should enhance AI's bias detection and mitigation to make it truly useful for diagnostic support.

NLPGuard: A Framework for Mitigating the Use of Protected Attributes by NLP Classifiers

AI regulations are expected to prohibit machine learning models from using sensitive attributes during training. However, the latest Natural Language Processing (NLP) classifiers, which rely on deep learning, operate as black-box systems, complicating the detection and remediation of such misuse. Traditional bias mitigation methods in NLP aim for comparable performance across different groups based on attributes like gender or race but fail to address the underlying issue of reliance on protected attributes. To partly fix that, we introduce NLPGuard, a framework for mitigating the reliance on protected attributes in NLP classifiers. NLPGuard takes an unlabeled dataset, an existing NLP classifier, and its training data as input, producing a modified training dataset that significantly reduces dependence on protected attributes without compromising accuracy. NLPGuard is applied to three classification tasks: identifying toxic language, sentiment analysis, and occupation classification. Our evaluation shows that current NLP classifiers heavily depend on protected attributes, with up to 23% of the most predictive words associated with these attributes. However, NLPGuard effectively reduces this reliance by up to 79%, while slightly improving accuracy.

From prediction to practice: mitigating bias and data shift in machine-learning models for chemotherapy-induced organ dysfunction across unseen cancers

ObjectivesRoutine monitoring of renal and hepatic function during chemotherapy ensures that treatment-related organ damage has not occurred and clearance of subsequent treatment is not hindered; however, frequency and timing are not optimal. Model bias and data heterogeneity concerns have hampered the ability of machine learning (ML) to be deployed into clinical practice. This study aims to develop models that could support individualised decisions on the timing of renal and hepatic monitoring while exploring the effect of data shift on model performance.Methods and analysisWe used retrospective data from three UK hospitals to develop and validate ML models predicting unacceptable rises in creatinine/bilirubin post cycle 3 for patients undergoing treatment for the following cancers: breast, colorectal, lung, ovarian and diffuse large B-cell lymphoma.ResultsWe extracted 3614 patients with no missing blood test data across cycles 1–6 of chemotherapy treatment. We improved on previous work by including predictions post cycle 3. Optimised for sensitivity, we achieve F2 scores of 0.7773 (bilirubin) and 0.6893 (creatinine) on unseen data. Performance is consistent on tumour types unseen during training (F2 bilirubin: 0.7423, F2 creatinine: 0.6820).ConclusionOur technique highlights the effectiveness of ML in clinical settings, demonstrating the potential to improve the delivery of care. Notably, our ML models can generalise to unseen tumour types. We propose gold-standard bias mitigation steps for ML models: evaluation on multisite data, thorough patient population analysis, and both formalised bias measures and model performance comparisons on patient subgroups. We demonstrate that data aggregation techniques have unintended consequences on model bias.

Wearing myopia glasses on GANs: Mitigating bias for pre-trained Generative Adversarial Networks via online prior perturbation

Pre-trained Generative Adversarial Networks (GANs) can provide rich information and make various downstream tasks beneficial. However, the training process of GANs potentially learns pre-existing biases in the historical data, which are not only inherited, but also amplified to some extent when using the generated data for prediction. Most available methods for bias mitigation utilize pre-processing or in-processing strategies to retrain GANs, which needs to access the raw data and also increases the time and computational cost required to train a fully expressive GAN. Therefore, we propose a novel post-processing mechanism to achieve fairness outputs of GANs through online prior perturbation. Unlike traditional output-based offline processing, our method mitigates biases from model inputs by designing a prior perturbation network, called “prior perturber”, to form a combined network with the pre-trained GAN. Specifically, we introduce a bias prediction network, called “bias predictor”, for online adversarial training with the prior perturber, which can decouple the generated representations from the bias features and improve the fairness of the downstream prediction results. In addition, the training process is independent of specific target labels, and the generated fair representations have good transferability. Experimental evaluations on five real datasets validate the effectiveness of the proposed method and the best utility is obtained on the same fairness level as compared to other output-based baseline methods. The result on fairness-utility trade-offs significantly improves the fairness of original samples by 56.40%, while the utility is reduced by only 2.60%.

Unveiling Market Sentiments: Finbert-Powered Analysis of Stock News Headlines

This document Sentiment analysis based on news and headlines is a big part of financial markets. Through the utilization of Hugging Face and FinBERT-a specialized model for financial sentiment analysis-and advanced natural language processing techniques, this study makes use of the flexibility of Hugging Face. This study concentrates on pre-processing techniques and the implementation of a model, emphasizing the critical evaluation and correction of inherent biases in sentiment analysis. Results of the experiment show that FinBERT is effective in addressing and reducing biases while extracting diverse sentiments from stock market headlines. This study emphasizes the importance of bias-conscious sentiment analysis for making more informed decisions in financial markets. It highlights the importance of advanced natural language processing models (NLP) like FinBERT and powerful frameworks like Hugging Face. Key Words: optics, Financial Sentiment Analysis, FinBERT, Hugging Face, Natural Language Processing (NLP), Bias Mitigation, Stock Market, Sentiment Analysis, Machine Learning, News Headlines, Investment Decision-making, Data Pre-processing, Neural Networks, Market Sentiments, Text Classification, NLP Frameworks

The Role of Machine Learning in Enhancing Corporate Financial Planning

Aim: To examine the role of machine learning in enhancing corporate financial planning. Problem Statement: In the past, corporate financial planning relied on statistical models and expert judgment which are often constrained by predefined assumptions and historical data. By so doing, they usually struggled to handle the non-linear and dynamic nature of financial markets causing limitations in their responsiveness and predictive power. Significance of Study: There is need to incorporate machine learning in corporate financial planning to provide a transformative technique via the introduction of advanced algorithms which are capable of analyzing diverse and vast datasets to uncover sophisticated relationships and patterns. Methodology: Recent relevant published articles in the area of machine learning in enhancing corporate financial planning were consulted. Relevant articles were sourced from the internet using the Google search engine. The abstract of the consulted published articles were thoroughly examined to study their relevance to the subject matter. Discussion: This review article examines the role of machine learning in enhancing corporate financial planning. It was found that machine learning techniques have crucial roles to play in the context of corporate financial planning via decision-making improvement, risk management, operational efficiency enhancement, fraud detection and enabling personalized experiences. Also, Machine Learning technologies have a wide variety of uses in corporate financial planning which enable banks and insurance companies in delivering tailored experiences that give their customers some preferences. Common identified machine learning approaches in corporate financial planning include predictive analytics, natural language processing, computer vision, reinforcement learning and anomaly detection approaches. Major areas of machine learning application include dynamic pricing and offers; regulatory compliance; personalized customer service; predictive analytics for customer retention; personalized product recommendations and so on. Furthermore, the implementation of Machine Learning (ML) in corporate financial planning and financial services industry provides many chances for enhancing customer experiences, improving efficiency and driving innovation. However, ML poses numerous challenges which should be addressed to attain the complete potential of these technologies. Conclusion: As technology continues to advance, financial institutions will need to route the challenges of bias mitigation, transparency and regulatory compliance while the ML full potential is equally leveraged. Also, machine learning has the potential to transform corporate financial planning via the provision of more adaptive, accurate and real-time insights.

Debiasing Judgements Using a Distributed Cognition Approach: A Scoping Review of Technological Strategies.

To review and synthesise research on technological debiasing strategies across domains, present a novel distributed cognition-based classification system, and discuss theoretical implications for the field. Distributed cognition theory is valuable for understanding and mitigating cognitive biases in high-stakes settings where sensemaking and problem-solving are contingent upon information representations and flows in the decision environment. Shifting the focus of debiasing from individuals to systems, technological debiasing strategies involve designing system components to minimise the negative impacts of cognitive bias on performance. To integrate these strategies into real-world practices effectively, it is imperative to clarify the current state of evidence and types of strategies utilised. We conducted systematic searches across six databases. Following screening and data charting, identified strategies were classified into (i) group composition and structure, (ii) information design and (iii) procedural debiasing, based on distributed cognition principles, and cognitive biases, classified into eight categories. Eighty articles met the inclusion criteria, addressing 100 debiasing investigations and 91 cognitive biases. A majority (80%) of the identified debiasing strategies were reportedly effective, whereas fourteen were ineffective and six were partially effective. Information design strategies were studied most, followed by procedural debiasing, and group structure and composition. Gaps and directions for future work are discussed. Through the lens of distributed cognition theory, technological debiasing represents a reconceptualisation of cognitive bias mitigation, showing promise for real-world application. The study results and debiasing classification presented can inform the design of high-stakes work systems to support cognition and minimise judgement errors.

The Adoption of Artificial Intelligence Technologies in Arab Newsrooms: Potentials and Challenges

Arab newsrooms and journalists are still attempting to process how to benefit from using artificial intelligence (AI) in their daily news production. Many have started experimenting with generative AI. However, in countries with strong economies, or in newsrooms funded by strong economies like in the UAE, Saudi Arabia, and Qatar, news organizations have endorsed AI technologies beyond ChatGPT and similar tools. The majority of Arab journalists, however, are yet to understand what is meant by AI. Very few news organizations in the region have strategically approached the question of AI. Arab newsrooms need to examine when and why they need to use AI technologies, whether they have the infrastructure or can afford the resources. Implementing AI in Arab newsrooms presents an additional set of challenges, including language complexities, cultural sensitivities, potential biases in algorithms, and the need for tailored solutions that resonate with local audiences. The ethical implications present its own set of challenges. ChatGPT has proven to generate inaccurate data in Arabic. Resources are to be put in place to ensure the model is trained on a diverse and comprehensive dataset of Arabic text. Implementing bias detection and mitigation techniques to ensure the models are free from biases and offensive content is crucial. Newsrooms operating in authoritarian regimes have to be aware of governments' attempts to use generative AI to whitewash and promote state-centered actions and policies. Maintaining editorial integrity remains an important consideration.

Mitigating Cognitive Bias to Improve Organizational Decisions: An Integrative Review, Framework, and Research Agenda

The detrimental influence of cognitive biases on decision-making and organizational performance is well established in management research. However, less attention has been given to bias mitigation interventions for improving organizational decisions. Drawing from the judgment and decision-making (JDM) literature, this paper offers a clear conceptualization of two approaches that mitigate bias via distinct cognitive mechanisms—debiasing and choice architecture—and presents a comprehensive integrative review of interventions tested experimentally within each approach. Observing a lack of comparative studies, we propose a novel framework that lays the foundation for future empirical research in bias mitigation. This framework identifies decision, organizational, and individual-level factors that are proposed to moderate the effectiveness of bias mitigation approaches across different contexts and can guide organizations in selecting the most suitable approach. By bridging JDM and management research, we offer a comprehensive research agenda and guidelines to select the most suitable evidence-based approach for improving decision-making processes and, ultimately, organizational performance.

Ethical and Legal Considerations in AI-Driven Health Cybersecurity

This article explores the ethical and legal challenges of implementing AI in health cybersecurity. It examines key issues including data privacy, informed consent, algorithmic bias, and balancing innovation with regulation. The rapid growth of AI in healthcare cybersecurity offers significant benefits but also introduces complex risks. The article analyzes data minimization, anonymization, bias mitigation, adaptive regulation, and international cooperation considerations. It presents data on the current landscape and proposes frameworks to address challenges. Overall, the article aims to provide a comprehensive overview of how to harness AI's potential for protecting health data while respecting patient rights and navigating evolving regulatory requirements.

Bias mitigation for fair automation of classification tasks

AbstractThe incorporation of machine learning algorithms into high‐risk decision‐making tasks has raised some alarms in the scientific community. Research shows that machine learning‐based technologies can contain biases that cause unfair decisions for certain population groups. The fundamental danger of ignoring this problem is that machine learning methods can not only reflect the biases present in our society but could also amplify them. This article presents the design and validation of a technology to assist the fair automation of classification problems. In essence, the proposal is based on taking advantage of the intermediate solutions generated during the resolution of classification problems through using Auto‐ML tools, in particular, AutoGOAL, to create unbiased/fair classifiers. The technology employs a multi‐objective optimization search to find the collection of models with the best trade‐offs between performance and fairness. To solve the optimization problem, we introduce a combination of Probabilistic Grammatical Evolution Search and NSGA‐II. The technology was evaluated using the Adult dataset from the UCI repository, a common benchmark in related research. Results were compared with other published results in scenarios with single and multiple fairness definitions. Our experiments demonstrate the technology's ability to automate classification tasks while incorporating fairness constraints. Additionally, our method achieves competitive results against other bias mitigation techniques. A notable advantage of our approach is its minimal requirement for machine learning expertise, thanks to its Auto‐ML foundation. This makes the technology accessible and valuable for advancing fairness in machine learning applications. The source code is available online for the research community.

Evaluation and mitigation of cognitive biases in medical language models.

Increasing interest in applying large language models (LLMs) to medicine is due in part to their impressive performance on medical exam questions. However, these exams do not capture the complexity of real patient-doctor interactions because of factors like patient compliance, experience, and cognitive bias. We hypothesized that LLMs would produce less accurate responses when faced with clinically biased questions as compared to unbiased ones. To test this, we developed the BiasMedQA dataset, which consists of 1273 USMLE questions modified to replicate common clinically relevant cognitive biases. We assessed six LLMs on BiasMedQA and found that GPT-4 stood out for its resilience to bias, in contrast to Llama 2 70B-chat and PMC Llama 13B, which showed large drops in performance. Additionally, we introduced three bias mitigation strategies, which improved but did not fully restore accuracy. Our findings highlight the need to improve LLMs' robustness to cognitive biases, in order to achieve more reliable applications of LLMs in healthcare.

Guidance for unbiased predictive information for healthcare decision-making and equity (GUIDE): considerations when race may be a prognostic factor.

Clinical prediction models (CPMs) are tools that compute the risk of an outcome given a set of patient characteristics and are routinely used to inform patients, guide treatment decision-making, and resource allocation. Although much hope has been placed on CPMs to mitigate human biases, CPMs may potentially contribute to racial disparities in decision-making and resource allocation. While some policymakers, professional organizations, and scholars have called for eliminating race as a variable from CPMs, others raise concerns that excluding race may exacerbate healthcare disparities and this controversy remains unresolved. The Guidance for Unbiased predictive Information for healthcare Decision-making and Equity (GUIDE) provides expert guidelines for model developers and health system administrators on the transparent use of race in CPMs and mitigation of algorithmic bias across contexts developed through a 5-round, modified Delphi process from a diverse 14-person technical expert panel (TEP). Deliberations affirmed that race is a social construct and that the goals of prediction are distinct from those of causal inference, and emphasized: the importance of decisional context (e.g., shared decision-making versus healthcare rationing); the conflicting nature of different anti-discrimination principles (e.g., anticlassification versus antisubordination principles); and the importance of identifying and balancing trade-offs in achieving equity-related goals with race-aware versus race-unaware CPMs for conditions where racial identity is prognostically informative. The GUIDE, comprising 31 key items in the development and use of CPMs in healthcare, outlines foundational principles, distinguishes between bias and fairness, and offers guidance for examining subgroup invalidity and using race as a variable in CPMs. This GUIDE presents a living document that supports appraisal and reporting of bias in CPMs to support best practice in CPM development and use.

AIOps in Action: Automating AI Deployment and Management of Large Language Models for Scalable and Ethical Operations

This comprehensive article explores the transformative role of Artificial Intelligence for IT Operations (AIOps) in the deployment and management of Large Language Models (LLMs). It delves into the automation strategies that streamline LLM deployment, including data preparation, model training optimization, and continuous integration and deployment practices. The article addresses the unique challenges in LLM management, such as resource allocation complexities and latency issues, presenting AIOps-driven solutions that leverage predictive analytics and dynamic scaling techniques. A significant focus is placed on the synergies between AIOps and MLOps, highlighting how their integration enhances model versioning, governance, and performance monitoring. The article also examines the critical aspects of real-time monitoring and incident management, showcasing how AIOps enables sophisticated anomaly detection and automated incident response. Ethical considerations in AIOps-driven LLM deployment are thoroughly discussed, emphasizing the importance of bias mitigation, transparency, and accountability. Looking ahead, the article explores future trends in AIOps for LLM management, including advancements in automation technologies and their implications for operational scalability and efficiency. Through a combination of theoretical analysis and practical case studies, this article provides a comprehensive overview of how AIOps is revolutionizing the landscape of AI operations, offering insights into both the current state and future potential of automated, scalable, and ethically responsible LLM management.