We study the challenging task of malware recognition on both known and novel unknown malware families, called malware open-set recognition (MOSR). Previous works usually assume the malware families are known to the classifier in a close-set scenario, i.e., testing families are the subset or at most identical to training families. However, novel unknown malware families frequently emerge in real-world applications, and as such, require to recognize malware instances in an open-set scenario, i.e., some unknown families are also included in the test-set, which has been rarely and non-thoroughly investigated in the cyber-security domain. One practical solution for MOSR may consider jointly classifying known and detecting unknown malware families by a single classifier (e.g., neural network) from the variance of the predicted probability distribution on known families. However, conventional well-trained classifiers usually tend to obtain overly high recognition probabilities in the outputs, especially when the instance feature distributions are similar to each other, e.g., unknown v.s. known malware families, and thus dramatically degrades the recognition on novel unknown malware families. In this paper, we propose a novel model that can conservatively synthesize malware instances to mimic unknown malware families and support a more robust training of the classifier. Moreover, we also build a new large-scale malware dataset, named MAL-100, to fill the gap of lacking large open-set malware benchmark dataset. Experimental results on two widely used malware datasets and our MAL-100 demonstrate the effectiveness of our model compared with other representative methods.

We employ Natural Language Processing techniques to analyse 377808 English song lyrics from the "Two Million Song Database" corpus, focusing on the expression of sexism across five decades (1960-2010) and the measurement of gender biases. Using a sexism classifier, we identify sexist lyrics at a larger scale than previous studies using small samples of manually annotated popular songs. Furthermore, we reveal gender biases by measuring associations in word embeddings learned on song lyrics. We find sexist content to increase across time, especially from male artists and for popular songs appearing in Billboard charts. Songs are also shown to contain different language biases depending on the gender of the performer, with male solo artist songs containing more and stronger biases. This is the first large scale analysis of this type, giving insights into language usage in such an influential part of popular culture.

Deep neural networks (DNNs) with a step-by-step introduction of inputs, which is constructed by imitating the somatosensory system in human body, known as SpinalNet have been implemented in this work on a Galaxy Zoo dataset. The input segmentation in SpinalNet has enabled the intermediate layers to take some of the inputs as well as output of preceding layers thereby reducing the amount of the collected weights in the intermediate layers. As a result of these, the authors of SpinalNet reported to have achieved in most of the DNNs they tested, not only a remarkable cut in the error but also in the large reduction of the computational costs. Having applied it to the Galaxy Zoo dataset, we are able to classify the different classes and/or sub-classes of the galaxies. Thus, we have obtained higher classification accuracies of 98.2, 95 and 82 percents between elliptical and spirals, between these two and irregulars, and between 10 sub-classes of galaxies, respectively.

Group fairness metrics are an established way of assessing the fairness of prediction-based decision-making systems. However, these metrics are still insufficiently linked to philosophical theories, and their moral meaning is often unclear. In this paper, we propose a comprehensive framework for group fairness metrics, which links them to more theories of distributive justice. The different group fairness metrics differ in their choices about how to measure the benefit or harm of a decision for the affected individuals, and what moral claims to benefits are assumed. Our unifying framework reveals the normative choices associated with standard group fairness metrics and allows an interpretation of their moral substance. In addition, this broader view provides a structure for the expansion of standard fairness metrics that we find in the literature. This expansion allows addressing several criticisms of standard group fairness metrics, specifically: (1) they are parity-based, i.e., they demand some form of equality between groups, which may sometimes be detrimental to marginalized groups; (2) they only compare decisions across groups but not the resulting consequences for these groups; and (3) the full breadth of the distributive justice literature is not sufficiently represented.

Although many fairness criteria have been proposed to ensure that machine learning algorithms do not exhibit or amplify our existing social biases, these algorithms are trained on datasets that can themselves be statistically biased. In this paper, we investigate the robustness of a number of existing (demographic) fairness criteria when the algorithm is trained on biased data. We consider two forms of dataset bias: errors by prior decision makers in the labeling process, and errors in measurement of the features of disadvantaged individuals. We analytically show that some constraints (such as Demographic Parity) can remain robust when facing certain statistical biases, while others (such as Equalized Odds) are significantly violated if trained on biased data. We also analyze the sensitivity of these criteria and the decision maker's utility to biases. We provide numerical experiments based on three real-world datasets (the FICO, Adult, and German credit score datasets) supporting our analytical findings. Our findings present an additional guideline for choosing among existing fairness criteria, or for proposing new criteria, when available datasets may be biased.

Malware open-set recognition (MOSR) aims at jointly classifying malware samples from known families and detect the ones from novel unknown families, respectively. Existing works mostly rely on a well-trained classifier considering the predicted probabilities of each known family with a threshold-based detection to achieve the MOSR. However, our observation reveals that the feature distributions of malware samples are extremely similar to each other even between known and unknown families. Thus the obtained classifier may produce overly high probabilities of testing unknown samples toward known families and degrade the model performance. In this paper, we propose the Multi-modal Dual-Embedding Networks, dubbed MDENet, to take advantage of comprehensive malware features (i.e., malware images and malware sentences) from different modalities to enhance the diversity of malware feature space, which is more representative and discriminative for down-stream recognition. Last, to further guarantee the open-set recognition, we dually embed the fused multi-modal representation into one primary space and an associated sub-space, i.e., discriminative and exclusive spaces, with contrastive sampling and rho-bounded enclosing sphere regularizations, which resort to classification and detection, respectively. Moreover, we also enrich our previously proposed large-scaled malware dataset MAL-100 with multi-modal characteristics and contribute an improved version dubbed MAL-100+. Experimental results on the widely used malware dataset Mailing and the proposed MAL-100+ demonstrate the effectiveness of our method.

Few-shot learning with sequence-processing neural networks (NNs) has recently attracted a new wave of attention in the context of large language models. In the standard N-way K-shot learning setting, an NN is explicitly optimised to learn to classify unlabelled inputs by observing a sequence of NK labelled examples. This pressures the NN to learn a learning algorithm that achieves optimal performance, given the limited number of training examples. Here we study an auxiliary loss that encourages further acceleration of few-shot learning, by applying recently proposed bootstrapped meta-learning to NN few-shot learners: we optimise the K-shot learner to match its own performance achievable by observing more than NK examples, using only NK examples. Promising results are obtained on the standard Mini-ImageNet dataset.

Safety and performance are key enablers for autonomous driving: on the one hand we want our autonomous vehicles (AVs) to be safe, while at the same time their performance (e.g., comfort or progression) is key to adoption. To effectively walk the tight-rope between safety and performance, AVs need to be risk-averse, but not entirely risk-avoidant. To facilitate safe-yet-performant driving, in this paper, we develop a task-aware risk estimator that assesses the risk a perception failure poses to the AV's motion plan. If the failure has no bearing on the safety of the AV's motion plan, then regardless of how egregious the perception failure is, our task-aware risk estimator considers the failure to have a low risk; on the other hand, if a seemingly benign perception failure severely impacts the motion plan, then our estimator considers it to have a high risk. In this paper, we propose a task-aware risk estimator to decide whether a safety maneuver needs to be triggered. To estimate the task-aware risk, first, we leverage the perception failure - detected by a perception monitor - to synthesize an alternative plausible model for the vehicle's surroundings. The risk due to the perception failure is then formalized as the "relative" risk to the AV's motion plan between the perceived and the alternative plausible scenario. We employ a statistical tool called copula, which models tail dependencies between distributions, to estimate this risk. The theoretical properties of the copula allow us to compute probably approximately correct (PAC) estimates of the risk. We evaluate our task-aware risk estimator using NuPlan and compare it with established baselines, showing that the proposed risk estimator achieves the best F1-score (doubling the score of the best baseline) and exhibits a good balance between recall and precision, i.e., a good balance of safety and performance.

Deep learning-based approaches have produced models with good insect classification accuracy; Most of these models are conducive for application in controlled environmental conditions. One of the primary emphasis of researchers is to implement identification and classification models in the real agriculture fields, which is challenging because input images that are wildly out of the distribution (e.g., images like vehicles, animals, humans, or a blurred image of an insect or insect class that is not yet trained on) can produce an incorrect insect classification. Out-of-distribution (OOD) detection algorithms provide an exciting avenue to overcome these challenge as it ensures that a model abstains from making incorrect classification prediction of non-insect and/or untrained insect class images. We generate and evaluate the performance of state-of-the-art OOD algorithms on insect detection classifiers. These algorithms represent a diversity of methods for addressing an OOD problem. Specifically, we focus on extrusive algorithms, i.e., algorithms that wrap around a well-trained classifier without the need for additional co-training. We compared three OOD detection algorithms: (i) Maximum Softmax Probability, which uses the softmax value as a confidence score, (ii) Mahalanobis distance-based algorithm, which uses a generative classification approach; and (iii) Energy-Based algorithm that maps the input data to a scalar value, called energy. We performed an extensive series of evaluations of these OOD algorithms across three performance axes: (a) \textit{Base model accuracy}: How does the accuracy of the classifier impact OOD performance? (b) How does the \textit{level of dissimilarity to the domain} impact OOD performance? and (c) \textit{Data imbalance}: How sensitive is OOD performance to the imbalance in per-class sample size?

In order to reliably process natural language, NLP systems must generalize to the long tail of rare utterances. We propose a method to create challenging benchmarks that require generalizing to the tail of the distribution by re-splitting existing datasets. We create 'Likelihood Splits' where examples that are assigned lower likelihood by a pre-trained language model (LM) are placed in the test set, and more likely examples are in the training set. This simple approach can be customized to construct meaningful train-test splits for a wide range of tasks. Likelihood Splits surface more challenges than random splits: relative error rates of state-of-the-art models increase by 59% for semantic parsing on Spider, 93% for natural language inference on SNLI, and 33% for yes/no question answering on BoolQ, on our splits compared with the corresponding random splits. Moreover, Likelihood Splits create fairer benchmarks than adversarial filtering; when the LM used to create the splits is also employed as the task model, our splits do not unfairly penalize the LM.