Most fair machine learning methods either highly rely on the sensitive information of the training samples or require a large modification on the target models, which hinders their practical application. To address this issue, we propose a two-stage training algorithm named FAIRIF. It minimizes the loss over the reweighted data set (second stage) where the sample weights are computed to balance the model performance across different demographic groups (first stage). FAIRIF can be applied on a wide range of models trained by stochastic gradient descent without changing the model, while only requiring group annotations on a small validation set to compute sample weights. Theoretically, we show that, in the classification setting, three notions of disparity among different groups can be mitigated by training with the weights. Experiments on synthetic data sets demonstrate that FAIRIF yields models with better fairness-utility trade-offs against various types of bias; and on real-world data sets, we show the effectiveness and scalability of FAIRIF. Moreover, as evidenced by the experiments with pretrained models, FAIRIF is able to alleviate the unfairness issue of pretrained models without hurting their performance.

You probably know at this point in the pandemic that if you test positive on a rapid test (and you've used the test correctly–you've got to swab pretty hard!) that the odds are overwhelming that you have COVID and are infectious. You have, no doubt, been warned that a negative result on a rapid test should be taken with a grain of salt, especially if you've been exposed to the virus or you have symptoms. And you might have heard two terms about tests thrown around: sensitivity and specificity, particularly if you've been trying to get a grip on exactly how accurate rapid tests are. As a mathematician, I think knowing a bit more about how sensitivity and specificity are calculated can help you better understand why a positive result should be headed--and a negative test should be read with some caution. The high specificity and rapid BinaxNOW antigen test turnaround time facilitate earlier isolation of infectious persons.

The other day my gf of 8 years got into it with my google assistant. I found out when she came into my home office and told me so. My gf told her off in some form. I pondered what her choice words may had been then lol'd loudly. Out loud i exclaimed "Ok, and what did google say".

Class imbalance can often degrade predictive performance of supervised learning algorithms. Balanced classes can be obtained by oversampling exact copies, with noise, or interpolation between nearest neighbours (as in traditional SMOTE methods). Oversampling tabular data using augmentation, as is typical in computer vision tasks, can be achieved with deep generative models. Deep generative models are effective data synthesisers due to their ability to capture complex underlying distributions. Synthetic data in healthcare can enhance interoperability between healthcare providers by ensuring patient privacy. Equipped with large synthetic datasets which do well to represent small patient groups, machine learning in healthcare can address the current challenges of bias and generalisability. This paper evaluates synthetic data generators ability to synthesise patient electronic health records. We test the utility of synthetic data for patient outcome classification, observing increased predictive performance when augmenting imbalanced datasets with synthetic data.

The tools and technology that are currently used to analyze chemical compound structures that identify polymer types in microplastics are not well-calibrated for environmentally weathered microplastics. Microplastics that have been degraded by environmental weathering factors can offer less analytic certainty than samples of microplastics that have not been exposed to weathering processes. Machine learning tools and techniques allow us to better calibrate the research tools for certainty in microplastics analysis. In this paper, we investigate whether the signatures (Raman shift values) are distinct enough such that well studied machine learning (ML) algorithms can learn to identify polymer types using a relatively small amount of labeled input data when the samples have not been impacted by environmental degradation. Several ML models were trained on a well-known repository, Spectral Libraries of Plastic Particles (SLOPP), that contain Raman shift and intensity results for a range of plastic particles, then tested on environmentally aged plastic particles (SloPP-E) consisting of 22 polymer types. After extensive preprocessing and augmentation, the trained random forest model was then tested on the SloPP-E dataset resulting in an improvement in classification accuracy of 93.81% from 89%.

The increasing automation in many areas of the Industry expressly demands to design efficient machine-learning solutions for the detection of abnormal events. With the ubiquitous deployment of sensors monitoring nearly continuously the health of complex infrastructures, anomaly detection can now rely on measurements sampled at a very high frequency, providing a very rich representation of the phenomenon under surveillance. In order to exploit fully the information thus collected, the observations cannot be treated as multivariate data anymore and a functional analysis approach is required. It is the purpose of this paper to investigate the performance of recent techniques for anomaly detection in the functional setup on real datasets. After an overview of the state-of-the-art and a visual-descriptive study, a variety of anomaly detection methods are compared. While taxonomies of abnormalities (e.g. shape, location) in the functional setup are documented in the literature, assigning a specific type to the identified anomalies appears to be a challenging task. Thus, strengths and weaknesses of the existing approaches are benchmarked in view of these highlighted types in a simulation study. Anomaly detection methods are next evaluated on two datasets, related to the monitoring of helicopters in flight and to the spectrometry of construction materials namely. The benchmark analysis is concluded by recommendation guidance for practitioners.

Terrorism is a major problem worldwide, causing thousands of fatalities and billions of dollars in damage every year. Toward the end of better understanding and mitigating these attacks, we present a set of machine learning models that learn from localized news data in order to predict whether a terrorist attack will occur on a given calendar date and in a given state. The best model--a Random Forest that learns from a novel variable-length moving average representation of the feature space--achieves area under the receiver operating characteristic scores $> .667$ on four of the five states that were impacted most by terrorism between 2015 and 2018. Our key findings include that modeling terrorism as a set of independent events, rather than as a continuous process, is a fruitful approach--especially when the events are sparse and dissimilar. Additionally, our results highlight the need for localized models that account for differences between locations. From a machine learning perspective, we found that the Random Forest model outperformed several deep models on our multimodal, noisy, and imbalanced data set, thus demonstrating the efficacy of our novel feature representation method in such a context. We also show that its predictions are relatively robust to time gaps between attacks and observed characteristics of the attacks. Finally, we analyze factors that limit model performance, which include a noisy feature space and small amount of available data. These contributions provide an important foundation for the use of machine learning in efforts against terrorism in the United States and beyond.

We present a novel method for controlling extrapolation in the prediction profiler in the JMP software. The prediction profiler is a graphical tool for exploring high dimensional prediction surfaces for statistical and machine learning models. The profiler contains interactive cross-sectional views, or profile traces, of the prediction surface of a model. Our method helps users avoid exploring predictions that should be considered extrapolation. It also performs optimization over a constrained factor region that avoids extrapolation using a genetic algorithm. In simulations and real world examples, we demonstrate how optimal factor settings without constraint in the profiler are frequently extrapolated, and how extrapolation control helps avoid these solutions with invalid factor settings that may not be useful to the user.

I propose kernel ridge regression estimators for long term causal inference, where a short term experimental data set containing randomized treatment and short term surrogates is fused with a long term observational data set containing short term surrogates and long term outcomes. I propose estimators of treatment effects, dose responses, and counterfactual distributions with closed form solutions in terms of kernel matrix operations. I allow covariates, treatment, and surrogates to be discrete or continuous, and low, high, or infinite dimensional. For long term treatment effects, I prove $\sqrt{n}$ consistency, Gaussian approximation, and semiparametric efficiency. For long term dose responses, I prove uniform consistency with finite sample rates. For long term counterfactual distributions, I prove convergence in distribution.

Despite recent progress made by self-supervised methods in representation learning with residual networks, they still underperform supervised learning on the ImageNet classification benchmark, limiting their applicability in performance-critical settings. Building on prior theoretical insights from Mitrovic et al., 2021, we propose ReLICv2 which combines an explicit invariance loss with a contrastive objective over a varied set of appropriately constructed data views. ReLICv2 achieves 77.1% top-1 classification accuracy on ImageNet using linear evaluation with a ResNet50 architecture and 80.6% with larger ResNet models, outperforming previous state-of-the-art self-supervised approaches by a wide margin. Most notably, ReLICv2 is the first representation learning method to consistently outperform the supervised baseline in a like-for-like comparison using a range of standard ResNet architectures. Finally we show that despite using ResNet encoders, ReLICv2 is comparable to state-of-the-art self-supervised vision transformers.