Search

In many healthcare settings, it is both critical to consider fairness when building analytical applications but also uniquely unacceptable to lower model performance for one group to match that of another (e.g. fairness cannot be achieved by lowering the diagnostic ability of a model for one group to match that of another and lose overall diagnostic power). Therefore a modeler needs to maximize model performance across groups as much as possible, often while maintaining a model's interpretability, which is a challenge for a number of reasons. In this paper we therefore suggest a new modeling framework, FAIR, to maximize performance across imbalanced groups, based on existing linear regression approaches already commonly used in healthcare settings. We propose a full linear interaction model between groups and all other covariates, paired with a weighting of samples by group size and independent regularization penalties for each group. This efficient approach overcomes many of the limitations in current approaches and manages to balance learning from other groups with tailoring prediction to the small focal group(s). FAIR has an added advantage in that it still allows for model interpretability in research and clinical settings. We demonstrate its usefulness with numerical and health data experiments., Comment: 12 pages, 4 figures

Objective. Demographic groups are often represented at different rates in medical datasets. These differences can create bias in machine learning algorithms, with higher levels of performance for better-represented groups. One promising solution to this problem is to generate synthetic data to mitigate potential adverse effects of non-representative data sets. Methods. We build on recent advances in LLM-based synthetic data generation to create a pipeline where the synthetic data is generated separately for each demographic group. We conduct our study using MIMIC-IV and Framingham "Offspring and OMNI-1 Cohorts" datasets. We prompt GPT4-Turbo to create group-specific data, providing training examples and the dataset context. An exploratory analysis is conducted to ascertain the quality of the generated data. We then evaluate the utility of the synthetic data for augmentation of a training dataset in a downstream machine learning task, focusing specifically on model performance metrics across groups. Results. The performance of GPT4-Turbo augmentation is generally superior but not always. In the majority of experiments our method outperforms standard modeling baselines, however, prompting GPT-4-Turbo to produce data specific to a group provides little to no additional benefit over a prompt that does not specify the group. Conclusion. We developed a method for using LLMs out-of-the-box to synthesize group-specific data to address imbalances in demographic representation in medical datasets. As another "tool in the toolbox", this method can improve model fairness and thus health equity. More research is needed to understand the conditions under which LLM generated synthetic data is useful for non-representative medical data sets., Comment: 26 pages, 4 figures

We present an analysis of the representation of gender as a data dimension in data visualizations and propose a set of considerations around visual variables and annotations for gender-related data. Gender is a common demographic dimension of data collected from study or survey participants, passengers, or customers, as well as across academic studies, especially in certain disciplines like sociology. Our work contributes to multiple ongoing discussions on the ethical implications of data visualizations. By choosing specific data, visual variables, and text labels, visualization designers may, inadvertently or not, perpetuate stereotypes and biases. Here, our goal is to start an evolving discussion on how to represent data on gender in data visualizations and raise awareness of the subtleties of choosing visual variables and words in gender visualizations. In order to ground this discussion, we collected and coded gender visualizations and their captions from five different scientific communities (Biology, Politics, Social Studies, Visualisation, and Human-Computer Interaction), in addition to images from Tableau Public and the Information Is Beautiful awards showcase. Overall we found that representation types are community-specific, color hue is the dominant visual channel for gender data, and nonconforming gender is under-represented. We end our paper with a discussion of considerations for gender visualization derived from our coding and the literature and recommendations for large data collection bodies. A free copy of this paper and all supplemental materials are available at https://osf.io/v9ams/, Comment: 11 page

This paper collects a set of open research questions on how to visualize sociodemographic data. Sociodemographic data is a common part of datasets related to people, including institutional censuses, health data systems, and human-resources fles. This data is sensitive, and its collection, sharing, and analysis require careful consideration. For instance, the European Union, through the General Data Protection Regulation (GDPR), protects the collection and processing of any personal data, including sexual orientation, ethnicity, and religion. Data visualization of sociodemographic data can reinforce stereotypes, marginalize groups, and lead to biased decision-making. It is, therefore, critical that these visualizations are created based on good, equitable design principles. In this paper, we discuss and provide a set of open research questions around the visualization of sociodemographic data. Our work contributes to an ongoing refection on representing data about people and highlights some important future research directions for the VIS community. A version of this paper and its fgures are available online at osf.io/a2u9c.

The objective of this paper is to explore how financial big data and machine learning methods can be applied to model and understand financial products. We focus on residential mortgage backed securities, resMBS, which were at the heart of the 2008 US financial crisis. These securities are contained within a prospectus and have a complex waterfall payoff structure. Multiple financial institutions form a supply chain to create prospectuses. To model this supply chain, we use unsupervised probabilistic methods, particularly dynamic topics models (DTM), to extract a set of features (topics) reflecting community formation and temporal evolution along the chain. We then provide insight into the performance of the resMBS securities and the impact of the supply chain through a series of increasingly comprehensive models. First, models at the security level directly identify salient features of resMBS securities that impact their performance. We then extend the model to include prospectus level features and demonstrate that the composition of the prospectus is significant. Our model also shows that communities along the supply chain that are associated with the generation of the prospectuses and securities have an impact on performance. We are the first to show that toxic communities that are closely linked to financial institutions that played a key role in the subprime crisis can increase the risk of failure of resMBS securities.