One-class classification is a challenging subfield of machine learning in which so-called data descriptors are used to predict membership of a class based solely on positive examples of that class, and no counter-examples. A number of data descriptors that have been shown to perform well in previous studies of one-class classification, like the Support Vector Machine (SVM), require setting one or more hyperparameters. There has been no systematic attempt to date to determine optimal default values for these hyperparameters, which limits their ease of use, especially in comparison with hyperparameter-free proposals like the Isolation Forest (IF). We address this issue by determining optimal default hyperparameter values across a collection of 246 one-class classification problems derived from 50 different real-world datasets. In addition, we propose a new data descriptor, Average Localised Proximity (ALP) to address certain issues with existing approaches based on nearest neighbour distances. Finally, we evaluate classification performance using a leave-one-dataset-out procedure, and find strong evidence that ALP outperforms IF and a number of other data descriptors, as well as weak evidence that it outperforms SVM, making ALP a good default choice.

Consider the classical supervised learning problem: we are given data $(y_i,{\boldsymbol x}_i)$, $i\le n$, with $y_i$ a response and ${\boldsymbol x}_i\in {\mathcal X}$ a covariates vector, and try to learn a model $f:{\mathcal X}\to{\mathbb R}$ to predict future responses. Random features methods map the covariates vector ${\boldsymbol x}_i$ to a point ${\boldsymbol \phi}({\boldsymbol x}_i)$ in a higher dimensional space ${\mathbb R}^N$, via a random featurization map ${\boldsymbol \phi}$. We study the use of random features methods in conjunction with ridge regression in the feature space ${\mathbb R}^N$. This can be viewed as a finite-dimensional approximation of kernel ridge regression (KRR), or as a stylized model for neural networks in the so called lazy training regime. We define a class of problems satisfying certain spectral conditions on the underlying kernels, and a hypercontractivity assumption on the associated eigenfunctions. These conditions are verified by classical high-dimensional examples. Under these conditions, we prove a sharp characterization of the error of random features ridge regression. In particular, we address two fundamental questions: $(1)$~What is the generalization error of KRR? $(2)$~How big $N$ should be for the random features approximation to achieve the same error as KRR? In this setting, we prove that KRR is well approximated by a projection onto the top $\ell$ eigenfunctions of the kernel, where $\ell$ depends on the sample size $n$. We show that the test error of random features ridge regression is dominated by its approximation error and is larger than the error of KRR as long as $N\le n^{1-\delta}$ for some $\delta>0$. We characterize this gap. For $N\ge n^{1+\delta}$, random features achieve the same error as the corresponding KRR, and further increasing $N$ does not lead to a significant change in test error.

Disagreements are pervasive in human communication. In this paper we investigate what makes disagreement constructive. To this end, we construct WikiDisputes, a corpus of 7 425 Wikipedia Talk page conversations that contain content disputes, and define the task of predicting whether disagreements will be escalated to mediation by a moderator. We evaluate feature-based models with linguistic markers from previous work, and demonstrate that their performance is improved by using features that capture changes in linguistic markers throughout the conversations, as opposed to averaged values. We develop a variety of neural models and show that taking into account the structure of the conversation improves predictive accuracy, exceeding that of feature-based models. We assess our best neural model in terms of both predictive accuracy and uncertainty by evaluating its behaviour when it is only exposed to the beginning of the conversation, finding that model accuracy improves and uncertainty reduces as models are exposed to more information.

Objectives: To evaluate the consequences of the framing of machine learning risk prediction models. We evaluate how framing affects model performance and model learning in four different approaches previously applied in published artificial-intelligence (AI) models. Setting and participants: We analysed structured secondary healthcare data from 221,283 citizens from four Danish municipalities who were 18 years of age or older. Results: The four models had similar population level performance (a mean area under the receiver operating characteristic curve of 0.73 to 0.82), in contrast to the mean average precision, which varied greatly from 0.007 to 0.385. Correspondingly, the percentage of missing values also varied between framing approaches. The on-clinical-demand framing, which involved samples for each time the clinicians made an early warning score assessment, showed the lowest percentage of missing values among the vital sign parameters, and this model was also able to learn more temporal dependencies than the others. The Shapley additive explanations demonstrated opposing interpretations of SpO2 in the prediction of sepsis as a consequence of differentially framed models. Conclusions: The profound consequences of framing mandate attention from clinicians and AI developers, as the understanding and reporting of framing are pivotal to the successful development and clinical implementation of future AI technology. Model framing must reflect the expected clinical environment. The importance of proper problem framing is by no means exclusive to sepsis prediction and applies to most clinical risk prediction models.

Explainable Artificial Intelligence (XAI) has in recent years become a well-suited framework to generate human understandable explanations of black box models. In this paper, we present a novel XAI visual explanation algorithm denoted SIDU that can effectively localize entire object regions responsible for prediction in a full extend. We analyze its robustness and effectiveness through various computational and human subject experiments. In particular, we assess the SIDU algorithm using three different types of evaluations (Application, Human and Functionally-Grounded) to demonstrate its superior performance. The robustness of SIDU is further studied in presence of adversarial attack on black box models to better understand its performance.

High-throughput sequencing technologies have rapidly developed during the past years and have become an essential tool in plant sciences. However, the analysis of genomic data remains challenging and relies mostly on the performance of automatic pipelines. Frequently applied pipelines involve the alignment of sequence reads against a reference sequence and the identification of sequence variants. Since most benchmarking studies of bioinformatics tools for this purpose have been conducted on human datasets, there is a lack of benchmarking studies in plant sciences. In this study, we evaluated the performance of 50 different variant calling pipelines, including five read mappers and ten variant callers, on six real plant datasets of the model organism Arabidopsis thaliana. Sets of variants were evaluated based on various parameters including sensitivity and specificity. We found that all investigated tools are suitable for analysis of NGS data in plant research. When looking at different performance metrics, BWA-MEM and Novoalign were the best mappers and GATK returned the best results in the variant calling step.

Controlling false discovery rate (FDR) while leveraging the side information of multiple hypothesis testing is an emerging research topic in modern data science. Existing methods rely on the test-level covariates while ignoring possible hierarchy among the covariates. This strategy may not be optimal for complex large-scale problems, where hierarchical information often exists among those test-level covariates. We propose NeurT-FDR which boosts statistical power and controls FDR for multiple hypothesis testing while leveraging the hierarchy among test-level covariates. Our method parametrizes the test-level covariates as a neural network and adjusts the feature hierarchy through a regression framework, which enables flexible handling of high-dimensional features as well as efficient end-to-end optimization. We show that NeurT-FDR has strong FDR guarantees and makes substantially more discoveries in synthetic and real datasets compared to competitive baselines.

The literature on using yield curves to forecast recessions typically measures the term spread as the difference between the 10-year and the three-month Treasury rates. Furthermore, using the term spread constrains the long- and short-term interest rates to have the same absolute effect on the recession probability. In this study, we adopt a machine learning method to investigate whether the predictive ability of interest rates can be improved. The machine learning algorithm identifies the best maturity pair, separating the effects of interest rates from those of the term spread. Our comprehensive empirical exercise shows that, despite the likelihood gain, the machine learning approach does not significantly improve the predictive accuracy, owing to the estimation error. Our finding supports the conventional use of the 10-year--three-month Treasury yield spread. This is robust to the forecasting horizon, control variable, sample period, and oversampling of the recession observations.

NLP impacts how firms provide products to users, content individuals receive through search and social media, and how While artificial intelligence provides the backbone for many tools individuals interact with news and emails. Despite the growing people use around the world, recent work has brought to attention importance of NLP algorithms in shaping our lives, recently scholars, that the algorithms powering AI are not free of politics, stereotypes, policymakers, and the business community have raised the and bias. While most work in this area has focused on the ways alarm of how gender and racial biases may be baked into these algorithms. in which AI can exacerbate existing inequalities and discrimination, Because they are trained on human data, the algorithms very little work has studied how governments actively shape themselves can replicate implicit and explicit human biases and training data. We describe how censorship has affected the development aggravate discrimination [6, 8, 39]. Additionally, training data that of Wikipedia corpuses, text data which are regularly used over-represents a subset of the population may do a worse job for pre-trained inputs into NLP algorithms. We show that word embeddings at predicting outcomes for other groups in the population [13].

There have been several research works proposing new Explainable AI (XAI) methods designed to generate model explanations having specific properties, or desiderata, such as fidelity, robustness, or human-interpretability. However, explanations are seldom evaluated based on their true practical impact on decision-making tasks. Without that assessment, explanations might be chosen that, in fact, hurt the overall performance of the combined system of ML model + end-users. This study aims to bridge this gap by proposing XAI Test, an application-grounded evaluation methodology tailored to isolate the impact of providing the end-user with different levels of information. We conducted an experiment following XAI Test to evaluate three popular post-hoc explanation methods -- LIME, SHAP, and TreeInterpreter -- on a real-world fraud detection task, with real data, a deployed ML model, and fraud analysts. During the experiment, we gradually increased the information provided to the fraud analysts in three stages: Data Only, i.e., just transaction data without access to model score nor explanations, Data + ML Model Score, and Data + ML Model Score + Explanations. Using strong statistical analysis, we show that, in general, these popular explainers have a worse impact than desired. Some of the conclusion highlights include: i) showing Data Only results in the highest decision accuracy and the slowest decision time among all variants tested, ii) all the explainers improve accuracy over the Data + ML Model Score variant but still result in lower accuracy when compared with Data Only; iii) LIME was the least preferred by users, probably due to its substantially lower variability of explanations from case to case.