Large language models (LLMs) have achieved huge success in numerous natural language process (NLP) tasks. However, it faces the challenge of significant resource consumption during inference. In this paper, we aim to improve the inference efficiency of LLMs by prompt caching, i.e., if the current prompt can be answered by the same response of a previous prompt, one can directly utilize that previous response without calling the LLM. Specifically, we focus on the prediction accuracy of prompt caching for single-round question-answering tasks via embedding similarity. The existing embeddings of prompts mostly focus on whether two prompts are semantically similar, which is not necessarily equivalent to whether the same response can answer them. Therefore, we propose a distillation-based method to fine-tune the existing embeddings for better caching prediction. Theoretically, we provide finite-sample guarantees for the convergence of our method under different types of loss functions. Empirically, we carefully construct a hard dataset based on Kwiatkowski et al. (2019) where the existing embedding model (Wang et al., 2022) only achieves an AUC of 0.51. We then fine-tune the above embedding model, which significantly improves the AUC of caching prediction from 0.51 to 0.81. We also conduct simulations demonstrating that our trained models achieve better caching efficiency than the previous embedding model.

In the realm of automatic speech recognition (ASR), the quest for models that not only perform with high accuracy but also offer transparency in their decision-making processes is crucial. The potential of quality estimation (QE) metrics is introduced and evaluated as a novel tool to enhance explainable artificial intelligence (XAI) in ASR systems. Through experiments and analyses, the capabilities of the NoRefER (No Reference Error Rate) metric are explored in identifying word-level errors to aid post-editors in refining ASR hypotheses. The investigation also extends to the utility of NoRefER in the corpus-building process, demonstrating its effectiveness in augmenting datasets with insightful annotations. The diagnostic aspects of NoRefER are examined, revealing its ability to provide valuable insights into model behaviors and decision patterns. This has proven beneficial for prioritizing hypotheses in post-editing workflows and fine-tuning ASR models. The findings suggest that NoRefER is not merely a tool for error detection but also a comprehensive framework for enhancing ASR systems' transparency, efficiency, and effectiveness. To ensure the reproducibility of the results, all source codes of this study are made publicly available.

AI fairness should not be considered a panacea: It may have the potential to make society more fair than ever, but it needs critical thought and outside help to make it happen.

Transfer learning (TL) has been demonstrated to improve DNN model performance when faced with a scarcity of training samples. However, the suitability of TL as a solution to reduce vulnerability of overfitted DNNs to privacy attacks is unexplored. A class of privacy attacks called membership inference attacks (MIAs) aim to determine whether a given sample belongs to the training dataset (member) or not (nonmember). We introduce Double-Dip, a systematic empirical study investigating the use of TL (Stage-1) combined with randomization (Stage-2) to thwart MIAs on overfitted DNNs without degrading classification accuracy. Our study examines the roles of shared feature space and parameter values between source and target models, number of frozen layers, and complexity of pretrained models. We evaluate Double-Dip on three (Target, Source) dataset paris: (i) (CIFAR-10, ImageNet), (ii) (GTSRB, ImageNet), (iii) (CelebA, VGGFace2). We consider four publicly available pretrained DNNs: (a) VGG-19, (b) ResNet-18, (c) Swin-T, and (d) FaceNet. Our experiments demonstrate that Stage-1 reduces adversary success while also significantly increasing classification accuracy of nonmembers against an adversary with either white-box or black-box DNN model access, attempting to carry out SOTA label-only MIAs. After Stage-2, success of an adversary carrying out a label-only MIA is further reduced to near 50%, bringing it closer to a random guess and showing the effectiveness of Double-Dip. Stage-2 of Double-Dip also achieves lower ASR and higher classification accuracy than regularization and differential privacy-based methods.

Anomaly detection plays a crucial role in ensuring network robustness. However, implementing intelligent alerting systems becomes a challenge when considering scenarios in which anomalies can be caused by both malicious and non-malicious events, leading to the difficulty of determining anomaly patterns. The lack of labeled data in the computer networking domain further exacerbates this issue, impeding the development of robust models capable of handling real-world scenarios. To address this challenge, in this paper, we propose an end-to-end anomaly detection model development pipeline. This framework makes it possible to consume user feedback and enable continuous user-centric model performance evaluation and optimization. We demonstrate the efficacy of the framework by way of introducing and bench-marking a new forecasting model -- named \emph{Lachesis} -- on a real-world networking problem. Experiments have demonstrated the robustness and effectiveness of the two proposed versions of \emph{Lachesis} compared with other models proposed in the literature. Our findings underscore the potential for improving the performance of data-driven products over their life cycles through a harmonized integration of user feedback and iterative development.

There has been much interest in recent years in learning good classifiers from data with noisy labels. Most work on learning from noisy labels has focused on standard loss-based performance measures. However, many machine learning problems require using non-decomposable performance measures which cannot be expressed as the expectation or sum of a loss on individual examples; these include for example the H-mean, Q-mean and G-mean in class imbalance settings, and the Micro $F_1$ in information retrieval. In this paper, we design algorithms to learn from noisy labels for two broad classes of multiclass non-decomposable performance measures, namely, monotonic convex and ratio-of-linear, which encompass all the above examples. Our work builds on the Frank-Wolfe and Bisection based methods of Narasimhan et al. (2015). In both cases, we develop noise-corrected versions of the algorithms under the widely studied family of class-conditional noise models. We provide regret (excess risk) bounds for our algorithms, establishing that even though they are trained on noisy data, they are Bayes consistent in the sense that their performance converges to the optimal performance w.r.t. the clean (non-noisy) distribution. Our experiments demonstrate the effectiveness of our algorithms in handling label noise.

A multi-modal machine learning system uses multiple unique data sources and types to improve its performance. This article proposes a system that combines results from several types of models, all of which are trained on different data signals. As an example to illustrate the efficacy of the system, an experiment is described in which multiple types of data are collected from rats suffering from seizures. This data includes electrocorticography readings, piezoelectric motion sensor data, and video recordings. Separate models are trained on each type of data, with the goal of classifying each time frame as either containing a seizure or not. After each model has generated its classification predictions, these results are combined. While each data signal works adequately on its own for prediction purposes, the significant imbalance in class labels leads to increased numbers of false positives, which can be filtered and removed by utilizing all data sources. This paper will demonstrate that, after postprocessing and combination techniques, classification accuracy is improved with this multi-modal system when compared to the performance of each individual data source.

In the high-stakes realm of healthcare, ensuring fairness in predictive models is crucial. Electronic Health Records (EHRs) have become integral to medical decision-making, yet existing methods for enhancing model fairness restrict themselves to unimodal data and fail to address the multifaceted social biases intertwined with demographic factors in EHRs. To mitigate these biases, we present FairEHR-CLP: a general framework for Fairness-aware Clinical Predictions with Contrastive Learning in EHRs. FairEHR-CLP operates through a two-stage process, utilizing patient demographics, longitudinal data, and clinical notes. First, synthetic counterparts are generated for each patient, allowing for diverse demographic identities while preserving essential health information. Second, fairness-aware predictions employ contrastive learning to align patient representations across sensitive attributes, jointly optimized with an MLP classifier with a softmax layer for clinical classification tasks. Acknowledging the unique challenges in EHRs, such as varying group sizes and class imbalance, we introduce a novel fairness metric to effectively measure error rate disparities across subgroups. Extensive experiments on three diverse EHR datasets on three tasks demonstrate the effectiveness of FairEHR-CLP in terms of fairness and utility compared with competitive baselines. FairEHR-CLP represents an advancement towards ensuring both accuracy and equity in predictive healthcare models.

Signal quality assessment (SQA) is required for monitoring the reliability of data acquisition systems, especially in AI-driven Predictive Maintenance (PMx) application contexts. SQA is vital for addressing "silent failures" of data acquisition hardware and software, which when unnoticed, misinform the users of data, creating the risk for incorrect decisions with unintended or even catastrophic consequences. We have developed an open-source software implementation of signal quality indices (SQIs) for the analysis of time-series data. We codify a range of SQIs, demonstrate them using established benchmark data, and show that they can be effective for signal quality assessment. We also study alternative approaches to denoising time-series data in an attempt to improve the quality of the already degraded signal, and evaluate them empirically on relevant real-world data. To our knowledge, our software toolkit is the first to provide an open source implementation of a broad range of signal quality assessment and improvement techniques validated on publicly available benchmark data for ease of reproducibility. The generality of our framework can be easily extended to assessing reliability of arbitrary time-series measurements in complex systems, especially when morphological patterns of the waveform shapes and signal periodicity are of key interest in downstream analyses.

We introduce a novel framework for incorporating human expertise into algorithmic predictions. Our approach focuses on the use of human judgment to distinguish inputs which `look the same' to any feasible predictive algorithm. We argue that this framing clarifies the problem of human/AI collaboration in prediction tasks, as experts often have access to information -- particularly subjective information -- which is not encoded in the algorithm's training data. We use this insight to develop a set of principled algorithms for selectively incorporating human feedback only when it improves the performance of any feasible predictor. We find empirically that although algorithms often outperform their human counterparts on average, human judgment can significantly improve algorithmic predictions on specific instances (which can be identified ex-ante). In an X-ray classification task, we find that this subset constitutes nearly 30% of the patient population. Our approach provides a natural way of uncovering this heterogeneity and thus enabling effective human-AI collaboration.