Fairness and bias correction in machine learning for depression prediction across four study populations.

A significant level of stigma and inequality exists in mental healthcare, especially in under-served populations. Inequalities are reflected in the data collected for scientific purposes. When not properly accounted for, machine learning (ML) models learned from data can reinforce these structural inequalities or biases. Here, we present a systematic study of bias in ML models designed to predict depression in four different case studies covering different countries and populations. We find that standard ML approaches regularly present biased behaviors. We also show that mitigation techniques, both standard and our own post-hoc method, can be effective in reducing the level of unfair bias. There is no one best ML model for depression prediction that provides equality of outcomes. This emphasizes the importance of analyzing fairness during model selection and transparent reporting about the impact of debiasing interventions. Finally, we also identify positive habits and open challenges that practitioners could follow to enhance fairness in their models.

Predicting ICU Mortality in Acute Respiratory Distress Syndrome Patients Using Machine Learning: The Predicting Outcome and STratifiCation of severity in ARDS (POSTCARDS) Study.

OBJECTIVES: To assess the value of machine learning approaches in the development of a multivariable model for early prediction of ICU death in patients with acute respiratory distress syndrome (ARDS). DESIGN: A development, testing, and external validation study using clinical data from four prospective, multicenter, observational cohorts. SETTING: A network of multidisciplinary ICUs. PATIENTS: A total of 1,303 patients with moderate-to-severe ARDS managed with lung-protective ventilation. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We developed and tested prediction models in 1,000 ARDS patients. We performed logistic regression analysis following variable selection by a genetic algorithm, random forest and extreme gradient boosting machine learning techniques. Potential predictors included demographics, comorbidities, ventilatory and oxygenation descriptors, and extrapulmonary organ failures. Risk modeling identified some major prognostic factors for ICU mortality, including age, cancer, immunosuppression, Pa o2 /F io2 , inspiratory plateau pressure, and number of extrapulmonary organ failures. Together, these characteristics contained most of the prognostic information in the first 24 hours to predict ICU mortality. Performance with machine learning methods was similar to logistic regression (area under the receiver operating characteristic curve [AUC], 0.87; 95% CI, 0.82-0.91). External validation in an independent cohort of 303 ARDS patients confirmed that the performance of the model was similar to a logistic regression model (AUC, 0.91; 95% CI, 0.87-0.94). CONCLUSIONS: Both machine learning and traditional methods lead to promising models to predict ICU death in moderate/severe ARDS patients. More research is needed to identify markers for severity beyond clinical determinants, such as demographics, comorbidities, lung mechanics, oxygenation, and extrapulmonary organ failure to guide patient management.

Modeling three sources of uncertainty in assisted reproductive technologies with probabilistic graphical models.

Embryo selection is a critical step in assisted reproduction: good selection criteria are expected to increase the probability of inducing a pregnancy. Machine learning techniques have been applied for implantation prediction or embryo quality assessment, which embryologists can use to make a decision about embryo selection. However, this is a highly uncertain real-world problem, and current proposals do not model always all the sources of uncertainty. We present a novel probabilistic graphical model that accounts for three different sources of uncertainty, the standard embryo and cycle viability, and a third one that represents any unknown factor that can drive a treatment to a failure in otherwise perfect conditions. We derive a parametric learning method based on the Expectation-Maximization strategy, which accounts for uncertainty issues. We empirically analyze the model within a real database consisting of 604 cycles (3125 embryos) carried out at Hospital Donostia (Spain). Embryologists followed the protocol of the Spanish Association for Reproduction Biology Studies (ASEBIR), based on morphological features, for embryo selection. Our model predictions are correlated with the ASEBIR protocol, which validates our model. The benefits of accounting for the different sources of uncertainty and the importance of the cycle characteristics are shown. Considering only transferred embryos, our model does not further discriminate them as implanted or failed, suggesting that the ASEBIR protocol could be understood as a thorough summary of the available morphological features.

A Robust Solution to Variational Importance Sampling of Minimum Variance.

Importance sampling is a Monte Carlo method where samples are obtained from an alternative proposal distribution. This can be used to focus the sampling process in the relevant parts of space, thus reducing the variance. Selecting the proposal that leads to the minimum variance can be formulated as an optimization problem and solved, for instance, by the use of a variational approach. Variational inference selects, from a given family, the distribution which minimizes the divergence to the distribution of interest. The Rényi projection of order 2 leads to the importance sampling estimator of minimum variance, but its computation is very costly. In this study with discrete distributions that factorize over probabilistic graphical models, we propose and evaluate an approximate projection method onto fully factored distributions. As a result of our evaluation it becomes apparent that a proposal distribution mixing the information projection with the approximate Rényi projection of order 2 could be interesting from a practical perspective.

Two datasets of defect reports labeled by a crowd of annotators of unknown reliability.

Classifying software defects according to any defined taxonomy is not straightforward. In order to be used for automatizing the classification of software defects, two sets of defect reports were collected from public issue tracking systems from two different real domains. Due to the lack of a domain expert, the collected defects were categorized by a set of annotators of unknown reliability according to their impact from IBM's orthogonal defect classification taxonomy. Both datasets are prepared to solve the defect classification problem by means of techniques of the learning from crowds paradigm (Hernández-González et al. [1]). Two versions of both datasets are publicly shared. In the first version, the raw data is given: the text description of defects together with the category assigned by each annotator. In the second version, the text of each defect has been transformed to a descriptive vector using text-mining techniques.

Fitting the data from embryo implantation prediction: Learning from label proportions.

Machine learning techniques have been previously used to assist clinicians to select embryos for human-assisted reproduction. This work aims to show how an appropriate modeling of the problem can contribute to improve machine learning techniques for embryo selection. In this study, a dataset of 330 consecutive cycles (and associated embryos) carried out by the Unit of Assisted Reproduction of the Hospital Donostia (Spain) throughout 18 months has been analyzed. The problem of the embryo selection has been modeled by a novel weakly supervised paradigm, learning from label proportions, which considers all the available data, including embryos whose fate cannot be certainly established. Furthermore, all the collected features, describing cycles and embryos, have been considered in a multi-variate data analysis. Our integral solution has been successfully tested. Experimental results show that the proposed technique consistently outperforms an equivalent approach based on standard supervised classification. Embryos in this study were selected for transference according to the criteria of the Spanish Association for Reproduction Biology Studies. Obtained classification models outperform these criteria, specifically reordering medium-quality embryos.