Research Papers

Artificial intelligence (AI) has demonstrated the ability to extract insights from data, but the issue of fairness remains a concern in high-stakes fields such as healthcare. Despite extensive discussion and efforts in algorithm development, AI fairness and clinical concerns have not been adequately addressed. In this paper, we discuss the misalignment between technical and clinical perspectives of AI fairness, highlight the barriers to AI fairness' translation to healthcare, advocate multidisciplinary collaboration to bridge the knowledge gap, and provide possible solutions to address the clinical concerns pertaining to AI fairness.

Counterfactual fairness is an approach to AI fairness that tries to make decisions based on the outcomes that an individual with some kind of sensitive status would have had without this status. This paper proposes Double Machine Learning (DML) Fairness which analogises this problem of counterfactual fairness in regression problems to that of estimating counterfactual outcomes in causal inference under the Potential Outcomes framework. It uses arbitrary machine learning methods to partial out the effect of sensitive variables on nonsensitive variables and outcomes. Assuming that the effects of the two sets of variables are additively separable, outcomes will be approximately equalised and individual-level outcomes will be counterfactually fair. This paper demonstrates the approach in a simulation study pertaining to discrimination in workplace hiring and an application on real data estimating the GPAs of law school students. It then discusses when it is appropriate to apply such a method to problems of real-world discrimination where constructs are conceptually complex and finally, whether DML Fairness can achieve justice in these settings.

Training dataset biases are by far the most scrutinized factors when explaining algorithmic biases of neural networks. In contrast, hyperparameters related to the neural network architecture have largely been ignored even though different network parameterizations are known to induce different implicit biases over learned features. For example, convolutional kernel size is known to affect the frequency content of features learned in CNNs. In this work, we present a causal framework for linking an architectural hyperparameter to out-of-distribution algorithmic bias. Our framework is experimental, in that we train several versions of a network with an intervention to a specific hyperparameter, and measure the resulting causal effect of this choice on performance bias when a particular out-of-distribution image perturbation is applied. In our experiments, we focused on measuring the causal relationship between convolutional kernel size and face analysis classification bias across different subpopulations (race/gender), with respect to high-frequency image details. We show that modifying kernel size, even in one layer of a CNN, changes the frequency content of learned features significantly across data subgroups leading to biased generalization performance even in the presence of a balanced dataset.

While the space of renewable energy forecasting has received significant attention in the last decade, literature has primarily focused on machine learning models that train on only one objective at a time. A host of classification (and regression) tasks in energy markets lead to highly imbalanced training data. Say, to balance reserves, it is natural for market regulators to have a choice to be more/less averse to false negatives (can lead to poor operating efficiency and costs) than to false positives (can lead to market shortfall). Besides accuracy, other metrics like algorithmic bias, RMBE (in regression problems), inferencing time, and model sparsity are also very crucial. This paper is amongst the firsts in the field of renewable energy forecasting that attempts to present a Pareto frontier of solutions (tradeoffs), that answers the question on handling multiple objectives by means of using the XGBoost model (Gradient Boosted Trees). Our proposed algorithm relies on using a sequence of weighted (uniform meshes) single objective model training routines. Real world data examples from the Amherst (Massachusetts, United States) solar energy prediction panels with both triobjective (focus on accuracy) and biojective (focus on fairness/bias) classification instances are considered. Numerical experiments appear promising and clear advantages over single objective methods are seen by observing the spread and variety of solutions (model configurations).

Equipping current decision-making tools with notions of fairness, equitability, or other ethically motivated outcomes, is one of the top priorities in recent research efforts in machine learning, AI, and optimization. In this paper, we investigate how to allocate limited resources to {locally interacting} communities in a way to maximize a pertinent notion of equitability. In particular, we look at the dynamic setting where the allocation is repeated across multiple periods (e.g., yearly), the local communities evolve in the meantime (driven by the provided allocation), and the allocations are modulated by feedback coming from the communities themselves. We employ recent mathematical tools stemming from data-driven feedback online optimization, by which communities can learn their (possibly unknown) evolution, satisfaction, as well as they can share information with the deciding bodies. We design dynamic policies that converge to an allocation that maximize equitability in the long term. We further demonstrate our model and methodology with realistic examples of healthcare and education subsidies design in Sub-Saharian countries. One of the key empirical takeaways from our setting is that long-term equitability is fragile, in the sense that it can be easily lost when deciding bodies weigh in other factors (e.g., equality in allocation) in the allocation strategy. Moreover, a naive compromise, while not providing significant advantage to the communities, can promote inequality in social outcomes.

Studies on supervised machine learning (ML) recommend involving workers from various backgrounds in training dataset labeling to reduce algorithmic bias. Moreover, sophisticated tasks for categorizing objects in images are necessary to improve ML performance, further complicating micro-tasks. This study aims to develop a task design incorporating the fair participation of people, regardless of their specific backgrounds or task's difficulty. By collaborating with 75 labelers from diverse backgrounds for 3 months, we analyzed workers' log-data and relevant narratives to identify the task's hurdles and helpers. The findings revealed that workers' decision-making tendencies varied depending on their backgrounds. We found that the community that positively helps workers and the machine's feedback perceived by workers could make people easily engaged in works. Hence, ML's bias could be expectedly mitigated. Based on these findings, we suggest an extended human-in-the-loop approach that connects labelers, machines, and communities rather than isolating individual workers.

Automatically understanding and recognising human affective states using images and computer vision can improve human-computer and human-robot interaction. However, privacy has become an issue of great concern, as the identities of people used to train affective models can be exposed in the process. For instance, malicious individuals could exploit images from users and assume their identities. In addition, affect recognition using images can lead to discriminatory and algorithmic bias, as certain information such as race, gender, and age could be assumed based on facial features. Possible solutions to protect the privacy of users and avoid misuse of their identities are to: (1) extract anonymised facial features, namely action units (AU) from a database of images, discard the images and use AUs for processing and training, and (2) federated learning (FL) i.e. process raw images in users' local machines (local processing) and send the locally trained models to the main processing machine for aggregation (central processing). In this paper, we propose a two-level deep learning architecture for affect recognition that uses AUs in level 1 and FL in level 2 to protect users' identities. The architecture consists of recurrent neural networks to capture the temporal relationships amongst the features and predict valence and arousal affective states. In our experiments, we evaluate the performance of our privacy-preserving architecture using different variations of recurrent neural networks on RECOLA, a comprehensive multimodal affective database. Our results show state-of-the-art performance of $0.426$ for valence and $0.401$ for arousal using the Concordance Correlation Coefficient evaluation metric, demonstrating the feasibility of developing models for affect recognition that are both accurate and ensure privacy.

Logical theories in the form of ontologies and similar artefacts in computing and IT are used for structuring, annotating, and querying data, among others, and therewith influence data analytics regarding what is fed into the algorithms. Algorithmic bias is a well-known notion, but what does bias mean in the context of ontologies that provide a structuring mechanism for an algorithm's input? What are the sources of bias there and how would they manifest themselves in ontologies? We examine and enumerate types of bias relevant for ontologies, and whether they are explicit or implicit. These eight types are illustrated with examples from extant production-level ontologies and samples from the literature. We then assessed three concurrently developed COVID-19 ontologies on bias and detected different subsets of types of bias in each one, to a greater or lesser extent. This first characterisation aims contribute to a sensitisation of ethical aspects of ontologies primarily regarding representation of information and knowledge.

Predicting Hyperkalemia in the ICU and Evaluation of Generalizability and Interpretability

Gloria Hyunjung Kwak, Christina Chen, Lowell Ling +3

Hyperkalemia is a potentially life-threatening condition that can lead to fatal arrhythmias. Early identification of high risk patients can inform clinical care to mitigate the risk. While hyperkalemia is often a complication of acute kidney injury (AKI), it also occurs in the absence of AKI. We developed predictive models to identify intensive care unit (ICU) patients at risk of developing hyperkalemia by using the Medical Information Mart for Intensive Care (MIMIC) and the eICU Collaborative Research Database (eICU-CRD). Our methodology focused on building multiple models, optimizing for interpretability through model selection, and simulating various clinical scenarios. In order to determine if our models perform accurately on patients with and without AKI, we evaluated the following clinical cases: (i) predicting hyperkalemia after AKI within 14 days of ICU admission, (ii) predicting hyperkalemia within 14 days of ICU admission regardless of AKI status, and compared different lead times for (i) and (ii). Both clinical scenarios were modeled using logistic regression (LR), random forest (RF), and XGBoost. Using observations from the first day in the ICU, our models were able to predict hyperkalemia with an AUC of (i) 0.79, 0.81, 0.81 and (ii) 0.81, 0.85, 0.85 for LR, RF, and XGBoost respectively. We found that 4 out of the top 5 features were consistent across the models. AKI stage was significant in the models that included all patients with or without AKI, but not in the models which only included patients with AKI. This suggests that while AKI is important for hyperkalemia, the specific stage of AKI may not be as important. Our findings require further investigation and confirmation.

Report prepared by the Montreal AI Ethics Institute In Response to Mila's Proposal for a Contact Tracing App

Allison Cohen, Abhishek Gupta

Contact tracing has grown in popularity as a promising solution to the COVID-19 pandemic. The benefits of automated contact tracing are two-fold. Contact tracing promises to reduce the number of infections by being able to: 1) systematically identify all of those that have been in contact with someone who has had COVID; and, 2) ensure those that have been exposed to the virus do not unknowingly infect others. "COVI" is the name of a recent contact tracing app developed by Mila and was proposed to help combat COVID-19 in Canada. The app was designed to inform each individual of their relative risk of being infected with the virus, which Mila claimed would empower citizens to make informed decisions about their movement and allow for a data-driven approach to public health policy; all the while ensuring data is safeguarded from governments, companies, and individuals. This article will provide a critical response to Mila's COVI White Paper. Specifically, this article will discuss: the extent to which diversity has been considered in the design of the app, assumptions surrounding users' interaction with the app and the app's utility, as well as unanswered questions surrounding transparency, accountability, and security. We see this as an opportunity to supplement the excellent risk analysis done by the COVI team to surface insights that can be applied to other contact- and proximity-tracing apps that are being developed and deployed across the world. Our hope is that, through a meaningful dialogue, we can ultimately help organizations develop better solutions that respect the fundamental rights and values of the communities these solutions are meant to serve.

Artificial intelligence (AI) has several transformative effects on healthcare, like to facilitate clinicians in disease detection and to manage patient care with enhanced precision. Due to the increasing complexity of medical data and the growing volume of diagnostic images, the substantial computational power needed to process such information creates both opportunities and challenges. To accomplish this objective, numerous methodologies have been proposed to integrate AI into disease detection systems. All these methodologies consider different aspects of clinical diagnosis, considering key parameters such as the type of AI model employed, the nature of input data, and the population demographics being analysed. In this paper, we review different state-of-the-art techniques which have been proposed for various areas of disease detection, specifically focusing on bias identification and mitigation. The proposed solutions are distinguished based on the type of input data they utilize, the architecture of model, and type of bias (algorithmic bias, demographic bias, model bias, and deployment bias). The time frame chosen to filter out relevant studies is between 2015 and 2026. Eventually, we review 23 related articles out of 64,389 analyzed from seven different digital libraries. The obtained results guide us to reach a conclusion with respect to state-of-the-art selected studies. Furthermore, these results reveal significant research gaps that can assist researchers in addressing AI-related bias in medical diagnosis. By prioritizing explainability, data diversity, and fairness in algorithmic design, as well as equitable access for affected patient populations, they can utilize available healthcare resources more effectively.

African Languages AI Safety Readiness Report

Israel Abebe Azime, Jesujoba O. Alabi, Tajuddeen Gwadabe +1

With the widespread adoption of large language models (LLMs) across languages and domains, AI safety has emerged as a critical global concern. However, due to resource constraints, research and development efforts tend to prioritize a limited number of high-resource languages, leaving many others, especially those spoken in Africa, underrepresented. While existing research on language "resource-ness" has largely examined whether sufficient resources are available to develop language technologies, the question of whether languages are adequately represented in AI safety research has received attention only more recently. This report examines the readiness of African languages for inclusion in AI safety discussions, drawing on existing literature and data. It considers four factors: the availability of AI-relevant resources, including text data; national AI policy support; performance on established text benchmarks; and speaker population. Based on these factors, languages are grouped into four readiness categories: fully ready, mostly ready, emerging, and foundational support needed. This classification is intended to inform the inclusion of African languages in AI safety benchmarks, highlight varying levels of preparedness, and guide strategic investment in data development, policy, and community engagement for underrepresented languages.