Artificial intelligence (AI) has revolutionized decision-making processes and systems throughout society and, in particular, has emerged as a significant technology in high-impact scenarios of national interest. Yet, despite AI's impressive predictive capabilities in controlled settings, it still suffers from a range of practical setbacks preventing its widespread use in various critical scenarios. In particular, it is generally unclear if a given AI system's predictions can be trusted by decision-makers in downstream applications. To address the need for more transparent, robust, and trustworthy AI systems, a suite of tools has been developed to quantify the uncertainty of AI predictions and, more generally, enable AI to "self-assess" the reliability of its predictions. In this manuscript, we categorize methods for AI self-assessment along several key dimensions and provide guidelines for selecting and designing the appropriate method for a practitioner's needs. In particular, we focus on uncertainty estimation techniques that consider the impact of self-assessment on the choices made by downstream decision-makers and on the resulting costs and benefits of decision outcomes. To demonstrate the utility of our methodology for self-assessment design, we illustrate its use for two realistic national-interest scenarios. This manuscript is a practical guide for machine learning engineers and AI system users to select the ideal self-assessment techniques for each problem.

Large language models (LLMs) are now being considered and even deployed for applications that support high-stakes decision-making, such as recruitment and clinical decisions. While several methods have been proposed for measuring bias, there remains a gap between predictions, which are what the proposed methods consider, and how they are used to make decisions. In this work, we introduce Rank-Allocational-Based Bias Index (RABBI), a model-agnostic bias measure that assesses potential allocational harms arising from biases in LLM predictions. We compare RABBI and current bias metrics on two allocation decision tasks. We evaluate their predictive validity across ten LLMs and utility for model selection. Our results reveal that commonly-used bias metrics based on average performance gap and distribution distance fail to reliably capture group disparities in allocation outcomes, whereas RABBI exhibits a strong correlation with allocation disparities. Our work highlights the need to account for how models are used in contexts with limited resource constraints.

Background: Ventilator-associated pneumonia (VAP) in traumatic brain injury (TBI) patients poses a significant mortality risk and imposes a considerable financial burden on patients and healthcare systems. Timely detection and prognostication of VAP in TBI patients are crucial to improve patient outcomes and alleviate the strain on healthcare resources. Methods: We implemented six machine learning models using the MIMIC-III database. Our methodology included preprocessing steps, such as feature selection with CatBoost and expert opinion, addressing class imbalance with the Synthetic Minority Oversampling Technique (SMOTE), and rigorous model tuning through 5-fold cross-validation to optimize hyperparameters. Key models evaluated included SVM, Logistic Regression, Random Forest, XGBoost, ANN, and AdaBoost. Additionally, we conducted SHAP analysis to determine feature importance and performed an ablation study to assess feature impacts on model performance. Results: XGBoost outperformed the baseline models and the best existing literature. We used metrics, including AUC, Accuracy, Specificity, Sensitivity, F1 Score, PPV, and NPV. XGBoost demonstrated the highest performance with an AUC of 0.940 and an Accuracy of 0.875, which are 23.4% and 23.5% higher than the best results in the existing literature, with an AUC of 0.706 and an Accuracy of 0.640, respectively. This enhanced performance underscores the models' effectiveness in clinical settings. Conclusions: This study enhances the predictive modeling of VAP in TBI patients, improving early detection and intervention potential. Refined feature selection and advanced ensemble techniques significantly boosted model accuracy and reliability, offering promising directions for future clinical applications and medical diagnostics research.

Machine learning has witnessed remarkable breakthroughs in recent years. As machine learning permeates various aspects of daily life, individuals and organizations increasingly interact with these systems, exhibiting a wide range of social and adversarial behaviors. These behaviors may have a notable impact on the behavior and performance of machine learning systems. Specifically, during these interactions, data may be generated by strategic individuals, collected by self-interested data collectors, possibly poisoned by adversarial attackers, and used to create predictors, models, and policies satisfying multiple objectives. As a result, the machine learning systems' outputs might degrade, such as the susceptibility of deep neural networks to adversarial examples (Shafahi et al., 2018; Szegedy et al., 2013) and the diminished performance of classic algorithms in the presence of strategic individuals (Ahmadi et al., 2021). Addressing these challenges is imperative for the success of machine learning in societal settings.

Background: Sepsis is a severe condition responsible for many deaths worldwide. Accurate prediction of sepsis outcomes is crucial for timely and effective treatment. Although previous studies have used ML to forecast outcomes, they faced limitations in feature selection and model comprehensibility, resulting in less effective predictions. Thus, this research aims to develop an interpretable and accurate ML model to help clinical professionals predict in-hospital mortality. Methods: We analyzed ICU patient records from the MIMIC-III database based on specific criteria and extracted relevant data. Our feature selection process included a literature review, clinical input refinement, and using Random Forest to select the top 35 features. We performed data preprocessing, including cleaning, imputation, standardization, and applied SMOTE for oversampling to address imbalance, resulting in 4,683 patients, with admission counts of 17,429. We compared the performance of Random Forest, Gradient Boosting, Logistic Regression, SVM, and KNN models. Results: The Random Forest model was the most effective in predicting sepsis-related in-hospital mortality. It outperformed other models, achieving an accuracy of 0.90 and an AUROC of 0.97, significantly better than the existing literature. Our meticulous feature selection contributed to the model's precision and identified critical determinants of sepsis mortality. These results underscore the pivotal role of data-driven ML in healthcare, especially for predicting in-hospital mortality due to sepsis. Conclusion: This study represents a significant advancement in predicting in-hospital sepsis mortality, highlighting the potential of ML in healthcare. The implications are profound, offering a data-driven approach that enhances decision-making in patient care and reduces in-hospital mortality.

Kernel ridge regression (KRR) is a popular class of machine learning models that has become an important tool for understanding deep learning. Much of the focus has been on studying the proportional asymptotic regime, $n \asymp d$, where $n$ is the number of training samples and $d$ is the dimension of the dataset. In this regime, under certain conditions on the data distribution, the kernel random matrix involved in KRR exhibits behavior akin to that of a linear kernel. In this work, we extend the study of kernel regression to the quadratic asymptotic regime, where $n \asymp d^2$. In this regime, we demonstrate that a broad class of inner-product kernels exhibit behavior similar to a quadratic kernel. Specifically, we establish an operator norm approximation bound for the difference between the original kernel random matrix and a quadratic kernel random matrix with additional correction terms compared to the Taylor expansion of the kernel functions. The approximation works for general data distributions under a Gaussian-moment-matching assumption with a covariance structure. This new approximation is utilized to obtain a limiting spectral distribution of the original kernel matrix and characterize the precise asymptotic training and generalization errors for KRR in the quadratic regime when $n/d^2$ converges to a non-zero constant. The generalization errors are obtained for both deterministic and random teacher models. Our proof techniques combine moment methods, Wick's formula, orthogonal polynomials, and resolvent analysis of random matrices with correlated entries.

Vision Language Models (VLMs) extend the capacity of LLMs to comprehensively understand vision information, achieving remarkable performance in many vision-centric tasks. Despite that, recent studies have shown that these models are susceptible to jailbreak attacks, which refer to an exploitative technique where malicious users can break the safety alignment of the target model and generate misleading and harmful answers. This potential threat is caused by both the inherent vulnerabilities of LLM and the larger attack scope introduced by vision input. To enhance the security of VLMs against jailbreak attacks, researchers have developed various defense techniques. However, these methods either require modifications to the model's internal structure or demand significant computational resources during the inference phase. Multimodal information is a double-edged sword. While it increases the risk of attacks, it also provides additional data that can enhance safeguards. Inspired by this, we propose $\underline{\textbf{C}}$ross-modality $\underline{\textbf{I}}$nformation $\underline{\textbf{DE}}$tecto$\underline{\textbf{R}}$ ($\textit{CIDER})$, a plug-and-play jailbreaking detector designed to identify maliciously perturbed image inputs, utilizing the cross-modal similarity between harmful queries and adversarial images. This simple yet effective cross-modality information detector, $\textit{CIDER}$, is independent of the target VLMs and requires less computation cost. Extensive experimental results demonstrate the effectiveness and efficiency of $\textit{CIDER}$, as well as its transferability to both white-box and black-box VLMs.

This work presents a novel framework for training Arabic nested embedding models through Matryoshka Embedding Learning, leveraging multilingual, Arabic-specific, and English-based models, to highlight the power of nested embeddings models in various Arabic NLP downstream tasks. Our innovative contribution includes the translation of various sentence similarity datasets into Arabic, enabling a comprehensive evaluation framework to compare these models across different dimensions. We trained several nested embedding models on the Arabic Natural Language Inference triplet dataset and assessed their performance using multiple evaluation metrics, including Pearson and Spearman correlations for cosine similarity, Manhattan distance, Euclidean distance, and dot product similarity. The results demonstrate the superior performance of the Matryoshka embedding models, particularly in capturing semantic nuances unique to the Arabic language. Results demonstrated that Arabic Matryoshka embedding models have superior performance in capturing semantic nuances unique to the Arabic language, significantly outperforming traditional models by up to 20-25\% across various similarity metrics. These results underscore the effectiveness of language-specific training and highlight the potential of Matryoshka models in enhancing semantic textual similarity tasks for Arabic NLP.

Empirical data can often be considered as samples from a set of probability distributions. Kernel methods have emerged as a natural approach for learning to classify these distributions. Although numerous kernels between distributions have been proposed, applying kernel methods to distribution regression tasks remains challenging, primarily because selecting a suitable kernel is not straightforward. Surprisingly, the question of learning a data-dependent distribution kernel has received little attention. In this paper, we propose a novel objective for the unsupervised learning of data-dependent distribution kernel, based on the principle of entropy maximization in the space of probability measure embeddings. We examine the theoretical properties of the latent embedding space induced by our objective, demonstrating that its geometric structure is well-suited for solving downstream discriminative tasks. Finally, we demonstrate the performance of the learned kernel across different modalities.

We evaluate the effectiveness of combining brain connectivity metrics with signal statistics for early stage Parkinson's Disease (PD) classification using electroencephalogram data (EEG). The data is from 5 arousal states - wakeful and four sleep stages (N1, N2, N3 and REM). Our pipeline uses an Ada Boost model for classification on a challenging early stage PD classification task with with only 30 participants (11 PD , 19 Healthy Control). Evaluating 9 brain connectivity metrics we find the best connectivity metric to be different for each arousal state with Phase Lag Index achieving the highest individual classification accuracy of 86\% on N1 data. Further to this our pipeline using regional signal statistics achieves an accuracy of 78\%, using brain connectivity only achieves an accuracy of 86\% whereas combining the two achieves a best accuracy of 91\%. This best performance is achieved on N1 data using Phase Lag Index (PLI) combined with statistics derived from the frequency characteristics of the EEG signal. This model also achieves a recall of 80 \% and precision of 96\%. Furthermore we find that on data from each arousal state, combining PLI with regional signal statistics improves classification accuracy versus using signal statistics or brain connectivity alone. Thus we conclude that combining brain connectivity statistics with regional EEG statistics is optimal for classifier performance on early stage Parkinson's. Additionally, we find outperformance of N1 EEG for classification of Parkinson's and expect this could be due to disrupted N1 sleep in PD. This should be explored in future work.