Neural networks have achieved impressive performance for data in the distribution which is the same as the training set but can produce an overconfident incorrect result for the data these networks have never seen. Therefore, it is essential to detect whether inputs come from out-of-distribution(OOD) in order to guarantee the safety of neural networks deployed in the real world. In this paper, we propose a simple and effective post-hoc technique, WeShort, to reduce the overconfidence of neural networks on OOD data. Our method is inspired by the observation of the internal residual structure, which shows the separation of the OOD and in-distribution (ID) data in the shortcut layer. Our method is compatible with different OOD detection scores and can generalize well to different architectures of networks. We demonstrate our method on various OOD datasets to show its competitive performances and provide reasonable hypotheses to explain why our method works. On the ImageNet benchmark, Weshort achieves state-of-the-art performance on the false positive rate (FPR95) and the area under the receiver operating characteristic (AUROC) on the family of post-hoc methods.

Given a learning problem with real-world tradeoffs, which cost function should the model be trained to optimize? This is the metric selection problem in machine learning. Despite its practical interest, there is limited formal guidance on how to select metrics for machine learning applications. This thesis outlines metric elicitation as a principled framework for selecting the performance metric that best reflects implicit user preferences. Once specified, the evaluation metric can be used to compare and train models. In this manuscript, we formalize the problem of Metric Elicitation and devise novel strategies for eliciting classification performance metrics using pairwise preference feedback over classifiers. Specifically, we provide novel strategies for eliciting linear and linear-fractional metrics for binary and multiclass classification problems, which are then extended to a framework that elicits group-fair performance metrics in the presence of multiple sensitive groups. All the elicitation strategies that we discuss are robust to both finite sample and feedback noise, thus are useful in practice for real-world applications. Using the tools and the geometric characterizations of the feasible confusion statistics sets from the binary, multiclass, and multiclass-multigroup classification setups, we further provide strategies to elicit from a wider range of complex, modern multiclass metrics defined by quadratic functions of confusion statistics by exploiting their local linear structure. From application perspective, we also propose to use the metric elicitation framework in optimizing complex black box metrics that is amenable to deep network training. Lastly, to bring theory closer to practice, we conduct a preliminary real-user study that shows the efficacy of the metric elicitation framework in recovering the users' preferred performance metric in a binary classification setup.

Open information extraction is an important NLP task that targets extracting structured information from unstructured text without limitations on the relation type or the domain of the text. This survey paper covers open information extraction technologies from 2007 to 2022 with a focus on new models not covered by previous surveys. We propose a new categorization method from the source of information perspective to accommodate the development of recent OIE technologies. In addition, we summarize three major approaches based on task settings as well as current popular datasets and model evaluation metrics. Given the comprehensive review, several future directions are shown from datasets, source of information, output form, method, and evaluation metric aspects.

The introduction of computational techniques to analyze chemical data has given rise to the analytical study of biological systems, known as "bioinformatics". One facet of bioinformatics is using machine learning (ML) technology to detect multivariable trends in various cases. Amongst the most pressing cases is predicting blood-brain barrier (BBB) permeability. The development of new drugs to treat central nervous system disorders presents unique challenges due to poor penetration efficacy across the blood-brain barrier. In this research, we aim to mitigate this problem through an ML model that analyzes chemical features. To do so: (i) An overview into the relevant biological systems and processes as well as the use case is given. (ii) Second, an in-depth literature review of existing computational techniques for detecting BBB permeability is undertaken. From there, an aspect unexplored across current techniques is identified and a solution is proposed. (iii) Lastly, a two-part in silico model to quantify likelihood of permeability of drugs with defined features across the BBB through passive diffusion is developed, tested, and reflected on. Testing and validation with the dataset determined the predictive logBB model's mean squared error to be around 0.112 units and the neuroinflammation model's mean squared error to be approximately 0.3 units, outperforming all relevant studies found.

Sharing medical data between institutions is difficult in practice due to data protection laws and official procedures within institutions. Therefore, most existing algorithms are trained on relatively small electroencephalogram (EEG) data sets which is likely to be detrimental to prediction accuracy. In this work, we simulate a case when the data can not be shared by splitting the publicly available data set into disjoint sets representing data in individual institutions. We propose to train a (local) detector in each institution and aggregate their individual predictions into one final prediction. Four aggregation schemes are compared, namely, the majority vote, the mean, the weighted mean and the Dawid-Skene method. The method was validated on an independent data set using only a subset of EEG channels. The ensemble reaches accuracy comparable to a single detector trained on all the data when sufficient amount of data is available in each institution. The weighted mean aggregation scheme showed best performance, it was only marginally outperformed by the Dawid--Skene method when local detectors approach performance of a single detector trained on all available data.

Clinical Text Notes (CTNs) contain physicians' reasoning process, written in an unstructured free text format, as they examine and interview patients. In recent years, several studies have been published that provide evidence for the utility of machine learning for predicting doctors' diagnoses from CTNs, a task known as ICD coding. Data annotation is time consuming, particularly when a degree of specialization is needed, as is the case for medical data. This paper presents a method of augmenting a sparsely annotated dataset of Icelandic CTNs with a machine-learned imputation in a semi-self-supervised manner. We train a neural network on a small set of annotated CTNs and use it to extract clinical features from a set of un-annotated CTNs. These clinical features consist of answers to about a thousand potential questions that a physician might find the answers to during a consultation of a patient. The features are then used to train a classifier for the diagnosis of certain types of diseases. We report the results of an evaluation of this data augmentation method over three tiers of data availability to the physician. Our data augmentation method shows a significant positive effect which is diminished when clinical features from the examination of the patient and diagnostics are made available. We recommend our method for augmenting scarce datasets for systems that take decisions based on clinical features that do not include examinations or tests.

Autonomous Systems (AS) are increasingly proposed, or used, in Safety Critical (SC) applications. Many such systems make use of sophisticated sensor suites and processing to provide scene understanding which informs the AS' decision-making. The sensor processing typically makes use of Machine Learning (ML) and has to work in challenging environments, further the ML-algorithms have known limitations,e.g., the possibility of false-negatives or false-positives in object classification. The well-established safety-analysis methods developed for conventional SC systems are not well-matched to AS, ML, or the sensing systems used by AS. This paper proposes an adaptation of well-established safety-analysis methods to address the specifics of perception-systems for AS, including addressing environmental effects and the potential failure-modes of ML, and provides a rationale for choosing particular sets of guidewords, or prompts, for safety-analysis. It goes on to show how the results of the analysis can be used to inform the design and verification of the AS and illustrates the new method by presenting a partial analysis of a road vehicle. Illustrations in the paper are primarily based on optical sensing, however the paper discusses the applicability of the method to other sensing modalities and its role in a wider safety process addressing the overall capabilities of AS.

This paper investigates whether socioeconomic factors are important for the hurricane performance of the electric power system in Florida. The investigation is performed using the Random Forest classifier with Mean Decrease of Accuracy (MDA) for measuring the importance of a set of factors that include hazard intensity, time to recovery from maximum impact, and socioeconomic characteristics of the affected population. The data set (at county scale) for this study includes socioeconomic variables from the 5-year American Community Survey (ACS), as well as wind velocities, and outage data of five hurricanes including Alberto and Michael in 2018, Dorian in 2019, and Eta and Isaias in 2020. The study shows that socioeconomic variables are considerably important for the system performance model. This indicates that social disparities may exist in the occurrence of power outages, which directly impact the resilience of communities and thus require immediate attention.

Email phishing has become more prevalent and grows more sophisticated over time. To combat this rise, many machine learning (ML) algorithms for detecting phishing emails have been developed. However, due to the limited email data sets on which these algorithms train, they are not adept at recognising varied attacks and, thus, suffer from concept drift; attackers can introduce small changes in the statistical characteristics of their emails or websites to successfully bypass detection. Over time, a gap develops between the reported accuracy from literature and the algorithm's actual effectiveness in the real world. This realises itself in frequent false positive and false negative classifications. To this end, we propose a multidimensional risk assessment of emails to reduce the feasibility of an attacker adapting their email and avoiding detection. This horizontal approach to email phishing detection profiles an incoming email on its main features. We develop a risk assessment framework that includes three models which analyse an email's (1) threat level, (2) cognitive manipulation, and (3) email type, which we combine to return the final risk assessment score. The Profiler does not require large data sets to train on to be effective and its analysis of varied email features reduces the impact of concept drift. Our Profiler can be used in conjunction with ML approaches, to reduce their misclassifications or as a labeller for large email data sets in the training stage. We evaluate the efficacy of the Profiler against a machine learning ensemble using state-of-the-art ML algorithms on a data set of 9000 legitimate and 900 phishing emails from a large Australian research organisation. Our results indicate that the Profiler's mitigates the impact of concept drift, and delivers 30% less false positive and 25% less false negative email classifications over the ML ensemble's approach.

Although many near-Earth objects have been found by ground-based telescopes, some fast-moving ones, especially those near detection limits, have been missed by observatories. We developed a convolutional neural network for detecting faint fast-moving near-Earth objects. It was trained with artificial streaks generated from simulations and was able to find these asteroid streaks with an accuracy of 98.7% and a false positive rate of 0.02% on simulated data. This program was used to search image data from the Zwicky Transient Facility (ZTF) in four nights in 2019, and it identified six previously undiscovered asteroids. The visual magnitudes of our detections range from ~19.0 - 20.3 and motion rates range from ~6.8 - 24 deg/day, which is very faint compared to other ZTF detections moving at similar motion rates. Our asteroids are also ~1 - 51 m diameter in size and ~5 - 60 lunar distances away at close approach, assuming their albedo values follow the albedo distribution function of known asteroids. The use of a purely simulated dataset to train our model enables the program to gain sensitivity in detecting faint and fast-moving objects while still being able to recover nearly all discoveries made by previously designed neural networks which used real detections to train neural networks. Our approach can be adopted by any observatory for detecting fast-moving asteroid streaks.