Search

Algorithmic vision, the computational process of making meaning from digital images or visual information, has changed the relationship between the image and the human subject. In this paper, I explicate on the role of algorithmic vision as a technique of algorithmic governance, the organization of a population by algorithmic means. With its roots in the United States post-war cybernetic sciences, the ontological status of the computational image undergoes a shift, giving way to the hegemonic use of automated facial recognition technologies towards predatory policing and profiling practices. By way of example, I argue that algorithmic vision reconfigures the philosophical links between vision, image, and truth, paradigmatically changing the way a human subject is represented through imagistic data. With algorithmic vision, the relationship between subject and representation challenges the humanistic discourse around images, calling for a critical displacement of the human subject from the center of an analysis of how computational images make meaning. I will explore the relationship between the operative image, the image that acts but is not seen by human eyes, and what Louise Amoore calls an "emergent subject," a subject that is made visible through algorithmic techniques (2013). Algorithmic vision reveals subjects to power in a mode that requires a new approach towards analyzing the entanglement and invisiblization of the human in automated decision-making systems. [ABSTRACT FROM AUTHOR]

Artificial intelligence models represented in machine learning algorithms are promising tools for risk assessment used to guide clinical and other health care decisions. Machine learning algorithms, however, may house biases that propagate stereotypes, inequities, and discrimination that contribute to socioeconomic health care disparities. The biases include those related to some sociodemographic characteristics such as race, ethnicity, gender, age, insurance, and socioeconomic status from the use of erroneous electronic health record data. Additionally, there is concern that training data and algorithmic biases in large language models pose potential drawbacks. These biases affect the lives and livelihoods of a significant percentage of the population in the United States and globally. The social and economic consequences of the associated backlash cannot be underestimated. Here, we outline some of the sociodemographic, training data, and algorithmic biases that undermine sound health care risk assessment and medical decision-making that should be addressed in the health care system. We present a perspective and overview of these biases by gender, race, ethnicity, age, historically marginalized communities, algorithmic bias, biased evaluations, implicit bias, selection/sampling bias, socioeconomic status biases, biased data distributions, cultural biases and insurance status bias, conformation bias, information bias and anchoring biases and make recommendations to improve large language model training data, including de-biasing techniques such as counterfactual role-reversed sentences during knowledge distillation, fine-tuning, prefix attachment at training time, the use of toxicity classifiers, retrieval augmented generation and algorithmic modification to mitigate the biases moving forward. [ABSTRACT FROM AUTHOR]

The topic of algorithmic fairness is of increasing importance to the Human–Computer Interaction research community following accumulating concerns regarding the use and deployment of Artificial Intelligence-based systems. How we conduct research on algorithmic fairness directly influences our inferences and conclusions regarding algorithmic fairness. To better understand the methodological decisions of studies focused on people's perceptions of algorithmic fairness, we systematic analysed relevant papers from the CHI and FAccT conferences. We identified 200 relevant papers published between 1993 and 2022 and assessed their study design, participant sample, and geographical location of participants and authors. Our results highlight that studies are predominantly cross-sectional, cover a wide range of participant roles, and that both authors and participants are primarily from the United States. Based on these findings, we reflect on the potential pitfalls and shortcomings in how the community studies algorithmic fairness. • Mapping of CHI and FAccT papers on algorithmic fairness published 1993–2022. • Over one-third of papers miss details on participant locality and compensation. • Algorithmic fairness research largely restricted to US and other Western countries. • Limited cross-country collaboration, most study samples from a single country. • Remote studies are most common, longitudinal studies are relatively rare. [ABSTRACT FROM AUTHOR]

Racial disparities in hospice care are well documented for patients with cancer, but the existence, direction, and extent of disparity findings are contradictory across the literature. Current methods to identify racial disparities aggregate data to produce single-value quality measures that exclude important patient quality elements and, consequently, lack information to identify actionable equity improvement insights. Our goal was to develop an explainable machine learning approach that elucidates healthcare disparities and provides more actionable quality improvement information. We infused clinical information with engineering systems modeling and data science to develop a time-by-utilization profile per patient group at each hospital using US Medicare hospice utilization data for a cohort of patients with advanced (poor-prognosis) cancer that died April-December 2016. We calculated the difference between group profiles for people of color and white people to identify racial disparity signatures. Using machine learning, we clustered racial disparity signatures across hospitals and compared these clusters to classic quality measures and hospital characteristics. With 45,125 patients across 362 hospitals, we identified 7 clusters; 4 clusters (n = 190 hospitals) showed more hospice utilization by people of color than white people, 2 clusters (n = 106) showed more hospice utilization by white people than people of color, and 1 cluster (n = 66) showed no difference. Within-hospital racial disparity behaviors cannot be predicted from quality measures, showing how the true shape of disparities can be distorted through the lens of quality measures. This approach elucidates the shape of hospice racial disparities algorithmically from the same data used to calculate quality measures. [ABSTRACT FROM AUTHOR]

Accurate forecasts can inform response to outbreaks. Most efforts in influenza forecasting have focused on predicting influenza-like activity, with fewer on influenza-related hospitalizations. We conducted a simulation study to evaluate a super learner's predictions of 3 seasonal measures of influenza hospitalizations in the United States: peak hospitalization rate, peak hospitalization week, and cumulative hospitalization rate. We trained an ensemble machine learning algorithm on 15,000 simulated hospitalization curves and generated weekly predictions. We compared the performance of the ensemble (weighted combination of predictions from multiple prediction algorithms), the best-performing individual prediction algorithm, and a naive prediction (median of a simulated outcome distribution). Ensemble predictions performed similarly to the naive predictions early in the season but consistently improved as the season progressed for all prediction targets. The best-performing prediction algorithm in each week typically had similar predictive accuracy compared with the ensemble, but the specific prediction algorithm selected varied by week. An ensemble super learner improved predictions of influenza-related hospitalizations, relative to a naive prediction. Future work should examine the super learner's performance using additional empirical data on influenza-related predictors (e.g. influenza-like illness). The algorithm should also be tailored to produce prospective probabilistic forecasts of selected prediction targets. [ABSTRACT FROM AUTHOR]

Merging satellite products and ground-based measurements is often required for obtaining precipitation datasets that simultaneously cover large regions with high density and are more accurate than pure satellite precipitation products. Machine and statistical learning regression algorithms are regularly utilized in this endeavor. At the same time, tree-based ensemble algorithms are adopted in various fields for solving regression problems with high accuracy and low computational costs. Still, information on which tree-based ensemble algorithm to select for correcting satellite precipitation products for the contiguous United States (US) at the daily time scale is missing from the literature. In this study, we worked towards filling this methodological gap by conducting an extensive comparison between three algorithms of the category of interest, specifically between random forests, gradient boosting machines (gbm) and extreme gradient boosting (XGBoost). We used daily data from the PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) and the IMERG (Integrated Multi-satellitE Retrievals for GPM) gridded datasets. We also used earth-observed precipitation data from the Global Historical Climatology Network daily (GHCNd) database. The experiments referred to the entire contiguous US and additionally included the application of the linear regression algorithm for benchmarking purposes. The results suggest that XGBoost is the best-performing tree-based ensemble algorithm among those compared. Indeed, the mean relative improvements that it provided with respect to linear regression (for the case that the latter algorithm was run with the same predictors as XGBoost) are equal to 52.66%, 56.26% and 64.55% (for three different predictor sets), while the respective values are 37.57%, 53.99% and 54.39% for random forests, and 34.72%, 47.99% and 62.61% for gbm. Lastly, the results suggest that IMERG is more useful than PERSIANN in the context investigated. [ABSTRACT FROM AUTHOR]

Early childhood is an important development period for establishing healthy physical activity (PA) habits. The objective of this study was to evaluate PA levels in a representative sample of U.S. preschool-aged children. The study sample included 301 participants (149 girls, 3–5 years of age) in the 2012 U.S. National Health and Examination Survey National Youth Fitness Survey. Participants were asked to wear an ActiGraph accelerometer on their wrist for 7 days. A machine learning random forest classification algorithm was applied to accelerometer data to estimate daily time spent in moderate- and vigorous-intensity PA (MVPA; the sum of minutes spent in running, walking, and other moderate- and vigorous-intensity PA) and total PA (the sum of MVPA and light-intensity PA). We estimated that U.S. preschool-aged children engaged in 28 min/day of MVPA and 361 min/day of total PA, on average. MVPA and total PA levels were not significantly different between males and females. This study revealed that U.S. preschool-aged children engage in lower levels of MVPA and higher levels of total PA than the minimum recommended by the World Health Organization. [ABSTRACT FROM AUTHOR]

Autism spectrum disorder (ASD) can be reliably diagnosed at 18 months, yet significant diagnostic delays persist in the United States. This double-blinded, multi-site, prospective, active comparator cohort study tested the accuracy of an artificial intelligence-based Software as a Medical Device designed to aid primary care healthcare providers (HCPs) in diagnosing ASD. The Device combines behavioral features from three distinct inputs (a caregiver questionnaire, analysis of two short home videos, and an HCP questionnaire) in a gradient boosted decision tree machine learning algorithm to produce either an ASD positive, ASD negative, or indeterminate output. This study compared Device outputs to diagnostic agreement by two or more independent specialists in a cohort of 18–72-month-olds with developmental delay concerns (425 study completers, 36% female, 29% ASD prevalence). Device output PPV for all study completers was 80.8% (95% confidence intervals (CI), 70.3%–88.8%) and NPV was 98.3% (90.6%–100%). For the 31.8% of participants who received a determinate output (ASD positive or negative) Device sensitivity was 98.4% (91.6%–100%) and specificity was 78.9% (67.6%–87.7%). The Device's indeterminate output acts as a risk control measure when inputs are insufficiently granular to make a determinate recommendation with confidence. If this risk control measure were removed, the sensitivity for all study completers would fall to 51.6% (63/122) (95% CI 42.4%, 60.8%), and specificity would fall to 18.5% (56/303) (95% CI 14.3%, 23.3%). Among participants for whom the Device abstained from providing a result, specialists identified that 91% had one or more complex neurodevelopmental disorders. No significant differences in Device performance were found across participants' sex, race/ethnicity, income, or education level. For nearly a third of this primary care sample, the Device enabled timely diagnostic evaluation with a high degree of accuracy. The Device shows promise to significantly increase the number of children able to be diagnosed with ASD in a primary care setting, potentially facilitating earlier intervention and more efficient use of specialist resources. [ABSTRACT FROM AUTHOR]

Electric water heaters represent 14% of the electricity consumption in residential buildings. An average household in the United States (U.S.) spends about USD 400–600 (0.45 ¢/L–0.68 ¢/L) on water heating every year. In this context, water heaters are often considered as a valuable asset for Demand Response (DR) and building energy management system (BEMS) applications. To this end, this study proposes a model-free deep reinforcement learning (RL) approach that aims to minimize the electricity cost of a water heater under a time-of-use (TOU) electricity pricing policy by only using standard DR commands. In this approach, a set of RL agents, with different look ahead periods, were trained using the deep Q-networks (DQN) algorithm and their performance was tested on an unseen pair of price and hot water usage profiles. The testing results showed that the RL agents can help save electricity cost in the range of 19% to 35% compared to the baseline operation without causing any discomfort to end users. Additionally, the RL agents outperformed rule-based and model predictive control (MPC)-based controllers and achieved comparable performance to optimization-based control. [ABSTRACT FROM AUTHOR]