Ensuring a neural network is not relying on protected attributes (e.g., race, sex, age) for prediction is crucial in advancing fair and trustworthy AI. While several promising methods for removing attribute bias in neural networks have been proposed, their limitations remain under-explored. To that end, in this work, we mathematically and empirically reveal the limitation of existing attribute bias removal methods in presence of strong bias and propose a new method that can mitigate this limitation. Specifically, we first derive a general non-vacuous information-theoretical upper bound on the performance of any attribute bias removal method in terms of the bias strength, revealing that they are effective only when the inherent bias in the dataset is relatively weak. Next, we derive a necessary condition for the existence of any method that can remove attribute bias regardless of the bias strength. Inspired by this condition, we then propose a new method using an adversarial objective that directly filters out protected attributes in the input space while maximally preserving all other attributes, without requiring any specific target label. The proposed method achieves state-of-the-art performance in both strong and moderate bias settings. We provide extensive experiments on synthetic, image, and census datasets, to verify the derived theoretical bound and its consequences in practice, and evaluate the effectiveness of the proposed method in removing strong attribute bias.

With a long history of traditional Graph Anomaly Detection (GAD) algorithms and recently popular Graph Neural Networks (GNNs), it is still not clear (1) how they perform under a standard comprehensive setting, (2) whether GNNs can outperform traditional algorithms such as tree ensembles, and (3) how about their efficiency on large-scale graphs. In response, we introduce GADBench -- a benchmark tool dedicated to supervised anomalous node detection in static graphs. GADBench facilitates a detailed comparison across 29 distinct models on ten real-world GAD datasets, encompassing thousands to millions ($\sim$6M) nodes. Our main finding is that tree ensembles with simple neighborhood aggregation can outperform the latest GNNs tailored for the GAD task. We shed light on the current progress of GAD, setting a robust groundwork for subsequent investigations in this domain. GADBench is open-sourced at https://github.com/squareRoot3/GADBench.

With the remarkable generation performance of large language models, ethical and legal concerns about using them have been raised, such as plagiarism and copyright issues. For such concerns, several approaches to watermark and detect LLM-generated text have been proposed very recently. However, we discover that the previous methods fail to function appropriately with code generation tasks because of the syntactic and semantic characteristics of code. Based on \citet{Kirchenbauer2023watermark}, we propose a new watermarking method, Selective WatErmarking via Entropy Thresholding (SWEET), that promotes "green" tokens only at the position with high entropy of the token distribution during generation, thereby preserving the correctness of the generated code. The watermarked code is detected by the statistical test and Z-score based on the entropy information. Our experiments on HumanEval and MBPP show that SWEET significantly improves the Pareto Frontier between the code correctness and watermark detection performance. We also show that notable post-hoc detection methods (e.g. DetectGPT) fail to work well in this task. Finally, we show that setting a reasonable entropy threshold is not much of a challenge. Code is available at https://github.com/hongcheki/sweet-watermark.

This paper introduces a novel approach for identifying the possible large language models (LLMs) involved in text generation. Instead of adding an additional classification layer to a base LM, we reframe the classification task as a next-token prediction task and directly fine-tune the base LM to perform it. We utilize the Text-to-Text Transfer Transformer (T5) model as the backbone for our experiments. We compared our approach to the more direct approach of utilizing hidden states for classification. Evaluation shows the exceptional performance of our method in the text classification task, highlighting its simplicity and efficiency. Furthermore, interpretability studies on the features extracted by our model reveal its ability to differentiate distinctive writing styles among various LLMs even in the absence of an explicit classifier. We also collected a dataset named OpenLLMText, containing approximately 340k text samples from human and LLMs, including GPT3.5, PaLM, LLaMA, and GPT2.

Uncertainty Quantification (UQ) has gained traction in an attempt to fix the black-box nature of Deep Learning. Specifically (medical) biosignals such as electroencephalography (EEG), electrocardiography (ECG), electroocculography (EOG) and electromyography (EMG) could benefit from good UQ, since these suffer from a poor signal to noise ratio, and good human interpretability is pivotal for medical applications and Brain Computer Interfaces. In this paper, we review the state of the art at the intersection of Uncertainty Quantification and Biosignal with Machine Learning. We present various methods, shortcomings, uncertainty measures and theoretical frameworks that currently exist in this application domain. Overall it can be concluded that promising UQ methods are available, but that research is needed on how people and systems may interact with an uncertainty model in a (clinical) environment.

A significant challenge for predictive maintenance in the pulp-and-paper industry is the infrequency of paper breaks during the production process. In this article, operational data is analyzed from a paper manufacturing machine in which paper breaks are relatively rare but have a high economic impact. Utilizing a dataset comprising 18,398 instances derived from a quality assurance protocol, we address the scarcity of break events (124 cases) that pose a challenge for machine learning predictive models. With the help of Conditional Generative Adversarial Networks (CTGAN) and Synthetic Minority Oversampling Technique (SMOTE), we implement a novel data augmentation framework. This method ensures that the synthetic data mirrors the distribution of the real operational data but also seeks to enhance the performance metrics of predictive modeling. Before and after the data augmentation, we evaluate three different machine learning algorithms-Decision Trees (DT), Random Forest (RF), and Logistic Regression (LR). Utilizing the CTGAN-enhanced dataset, our study achieved significant improvements in predictive maintenance performance metrics. The efficacy of CTGAN in addressing data scarcity was evident, with the models' detection of machine breaks (Class 1) improving by over 30% for Decision Trees, 20% for Random Forest, and nearly 90% for Logistic Regression. With this methodological advancement, this study contributes to industrial quality control and maintenance scheduling by addressing rare event prediction in manufacturing processes.

As more and more infection-specific machine learning models are developed and planned for clinical deployment, simultaneously running predictions from different models may provide overlapping or even conflicting information. It is important to understand the concordance and behavior of parallel models in deployment. In this study, we focus on two models for the early detection of hospital-acquired infections (HAIs): 1) the Infection Risk Index (IRI) and 2) the Ventilator-Associated Pneumonia (VAP) prediction model. The IRI model was built to predict all HAIs, whereas the VAP model identifies patients at risk of developing ventilator-associated pneumonia. These models could make important improvements in patient outcomes and hospital management of infections through early detection of infections and in turn, enable early interventions. The two models vary in terms of infection label definition, cohort selection, and prediction schema. In this work, we present a comparative analysis between the two models to characterize concordances and confusions in predicting HAIs by these models. The learnings from this study will provide important findings for how to deploy multiple concurrent disease-specific models in the future.

In the last decade, the United States has lost more than 500,000 people from an overdose involving prescription and illicit opioids (https://www.cdc.gov/drugoverdose/epidemic/index.html) making it a national public health emergency (USDHHS, 2017). To more effectively prevent unintentional opioid overdoses, medical practitioners require robust and timely tools that can effectively identify at-risk patients. Community-based social media platforms such as Reddit allow self-disclosure for users to discuss otherwise sensitive drug-related behaviors, often acting as indicators for opioid use disorder. Towards this, we present a moderate size corpus of 2500 opioid-related posts from various subreddits spanning 6 different phases of opioid use: Medical Use, Misuse, Addiction, Recovery, Relapse, Not Using. For every post, we annotate span-level extractive explanations and crucially study their role both in annotation quality and model development. We evaluate several state-of-the-art models in a supervised, few-shot, or zero-shot setting. Experimental results and error analysis show that identifying the phases of opioid use disorder is highly contextual and challenging. However, we find that using explanations during modeling leads to a significant boost in classification accuracy demonstrating their beneficial role in a high-stakes domain such as studying the opioid use disorder continuum. The dataset will be made available for research on Github in the formal version.

Uncertainty quantification in machine learning is a timely and vast field of research. In supervised learning, uncertainty can already occur in the very first stage of the training process, the labelling step. In particular, this is the case when not every instance can be unambiguously classified. The problem occurs for classifying instances, where classes may overlap or instances can not be clearly categorised. In other words, there is inevitable ambiguity in the annotation step and not necessarily a 'ground truth'. We look exemplary at the classification of satellite images. Each image is annotated independently by multiple labellers and classified into local climate zones (LCZs). For each instance we have multiple votes, leading to a distribution of labels rather than a single value. The main idea of this work is that we do not assume a ground truth label but embed the votes into a K-dimensional space, with K as the number of possible categories. The embedding is derived from the voting distribution in a Bayesian setup, modelled via a Dirichlet-Multinomial model. We estimate the model and posteriors using a stochastic Expectation Maximisation algorithm with Markov Chain Monte Carlo steps. While we focus on the particular example of LCZ classification, the methods developed in this paper readily extend to other situations where multiple annotators independently label texts or images. We also apply our approach to two other benchmark datasets for image classification to demonstrate this. Besides the embeddings themselves, we can investigate the resulting correlation matrices, which can be seen as generalised confusion matrices and reflect the semantic similarities of the original classes very well for all three exemplary datasets. The insights gained are valuable and can serve as general label embedding if a single ground truth per observation cannot be guaranteed.

Approaches to recommendation are typically evaluated in one of two ways: (1) via a (simulated) online experiment, often seen as the gold standard, or (2) via some offline evaluation procedure, where the goal is to approximate the outcome of an online experiment. Several offline evaluation metrics have been adopted in the literature, inspired by ranking metrics prevalent in the field of Information Retrieval. (Normalised) Discounted Cumulative Gain (nDCG) is one such metric that has seen widespread adoption in empirical studies, and higher (n)DCG values have been used to present new methods as the state-of-the-art in top-$n$ recommendation for many years. Our work takes a critical look at this approach, and investigates when we can expect such metrics to approximate the gold standard outcome of an online experiment. We formally present the assumptions that are necessary to consider DCG an unbiased estimator of online reward and provide a derivation for this metric from first principles, highlighting where we deviate from its traditional uses in IR. Importantly, we show that normalising the metric renders it inconsistent, in that even when DCG is unbiased, ranking competing methods by their normalised DCG can invert their relative order. Through a correlation analysis between off- and on-line experiments conducted on a large-scale recommendation platform, we show that our unbiased DCG estimates strongly correlate with online reward, even when some of the metric's inherent assumptions are violated. This statement no longer holds for its normalised variant, suggesting that nDCG's practical utility may be limited.