State-of-the-art automated machine learning systems for tabular data often employ cross-validation; ensuring that measured performances generalize to unseen data, or that subsequent ensembling does not overfit. However, using k-fold cross-validation instead of holdout validation drastically increases the computational cost of validating a single configuration. While ensuring better generalization and, by extension, better performance, the additional cost is often prohibitive for effective model selection within a time budget. We aim to make model selection with cross-validation more effective. Therefore, we study early stopping the process of cross-validation during model selection. We investigate the impact of early stopping on random search for two algorithms, MLP and random forest, across 36 classification datasets. We further analyze the impact of the number of folds by considering 3-, 5-, and 10-folds. In addition, we investigate the impact of early stopping with Bayesian optimization instead of random search and also repeated cross-validation. Our exploratory study shows that even a simple-to-understand and easy-to-implement method consistently allows model selection to converge faster; in ~94% of all datasets, on average by ~214%. Moreover, stopping cross-validation enables model selection to explore the search space more exhaustively by considering +167% configurations on average within one hour, while also obtaining better overall performance.

In lossy image compression, the objective is to achieve minimal signal distortion while compressing images to a specified bit rate. The increasing demand for visual analysis applications, particularly in classification tasks, has emphasized the significance of considering semantic distortion in compressed images. To bridge the gap between image compression and visual analysis, we propose a Rate-Distortion-Classification (RDC) model for lossy image compression, offering a unified framework to optimize the trade-off between rate, distortion, and classification accuracy. The RDC model is extensively analyzed both statistically on a multi-distribution source and experimentally on the widely used MNIST dataset. The findings reveal that the RDC model exhibits desirable properties, including monotonic non-increasing and convex functions, under certain conditions. This work provides insights into the development of human-machine friendly compression methods and Video Coding for Machine (VCM) approaches, paving the way for end-to-end image compression techniques in real-world applications.

Chronic Obstructive Pulmonary Disease (COPD) is a chronic inflammatory lung condition that causes airflow obstruction. The existing methods can only detect patients who already have COPD based on obvious features shown in the spirogram (In this article, the spirogram specifically involves measuring Volume-Flow curve time series). Early prediction of COPD risk is vital for monitoring COPD disease progression, slowing it down, or even preventing its onset. However, these methods fail to early predict an individual's probability of COPD in the future based on subtle features in the spirogram. To address this gap, for the first time, we propose DeepSpiro, a method based on deep learning for early prediction of future COPD risk. DeepSpiro consists of four parts. First, we construct Volume-Flow curves guided by Time-Volume instability smoothing (SpiroSmoother) to enhance the stability of the original Volume-Flow curves precisely. Second, we extract critical features from the evolution of varied-length key patches (SpiroEncoder) to capture the key temporal evolution from original high-dimensional dynamic sequences to a unified low-dimensional temporal representation. Third, we explain the model based on temporal attention and heterogeneous feature fusion (SpiroExplainer), which integrates information from heterogeneous data such as spirogram and demographic information. Fourth, we predict the risk of COPD based on the evolution of key patch concavity (SpiroPredictor), enabling accurate prediction of the risk of disease in high-risk patients who are not yet diagnosed, for up to 1, 2, 3, 4, 5 years, and beyond. We conduct experiments on the UK Biobank dataset. Results show that DeepSpiro achieves an AUC value of 0.8328 in the task of detecting COPD. In early prediction tasks, high-risk and low-risk groups show significant differences in the future, with a p-value of <0.001.

In this paper, we study the partial multi-label (PML) image classification problem, where each image is annotated with a candidate label set consists of multiple relevant labels and other noisy labels. Existing PML methods typically design a disambiguation strategy to filter out noisy labels by utilizing prior knowledge with extra assumptions, which unfortunately is unavailable in many real tasks. Furthermore, because the objective function for disambiguation is usually elaborately designed on the whole training set, it can be hardly optimized in a deep model with SGD on mini-batches. In this paper, for the first time we propose a deep model for PML to enhance the representation and discrimination ability. On one hand, we propose a novel curriculum based disambiguation strategy to progressively identify ground-truth labels by incorporating the varied difficulties of different classes. On the other hand, a consistency regularization is introduced for model retraining to balance fitting identified easy labels and exploiting potential relevant labels. Extensive experimental results on the commonly used benchmark datasets show the proposed method significantly outperforms the SOTA methods.

The widespread adoption of big data has ushered in a new era of data-driven decision-making, transforming numerous industries and sectors. However, the efficacy of these decisions hinges on the quality of the underlying data. Poor data quality can result in inaccurate analyses and deceptive conclusions. Managing the vast volume, velocity, and variety of data sources presents significant challenges, heightening the importance of addressing big data quality issues. While there has been increased attention from both academia and industry, current approaches often lack comprehensiveness and universality. They tend to focus on limited metrics, neglecting other dimensions of data quality. Moreover, existing methods are often context-specific, limiting their applicability across different domains. There is a clear need for intelligent, automated approaches leveraging artificial intelligence (AI) for advanced data quality corrections. To bridge these gaps, this Ph.D. thesis proposes a novel set of interconnected frameworks aimed at enhancing big data quality comprehensively. Firstly, we introduce new quality metrics and a weighted scoring system for precise data quality assessment. Secondly, we present a generic framework for detecting various quality anomalies using AI models. Thirdly, we propose an innovative framework for correcting detected anomalies through predictive modeling. Additionally, we address metadata quality enhancement within big data ecosystems. These frameworks are rigorously tested on diverse datasets, demonstrating their efficacy in improving big data quality. Finally, the thesis concludes with insights and suggestions for future research directions.

Data Shapley provides a principled approach to data valuation and plays a crucial role in data-centric machine learning (ML) research. Data selection is considered a standard application of Data Shapley. However, its data selection performance has shown to be inconsistent across settings in the literature. This study aims to deepen our understanding of this phenomenon. We introduce a hypothesis testing framework and show that Data Shapley's performance can be no better than random selection without specific constraints on utility functions. We identify a class of utility functions, monotonically transformed modular functions, within which Data Shapley optimally selects data. Based on this insight, we propose a heuristic for predicting Data Shapley's effectiveness in data selection tasks. Our experiments corroborate these findings, adding new insights into when Data Shapley may or may not succeed.

Anomaly detection, a critical facet in data analysis, involves identifying patterns that deviate from expected behavior. This research addresses the complexities inherent in anomaly detection, exploring challenges and adapting to sophisticated malicious activities. With applications spanning cybersecurity, healthcare, finance, and surveillance, anomalies often signify critical information or potential threats. Inspired by the success of Anomaly Generative Adversarial Network (AnoGAN) in image domains, our research extends its principles to tabular data. Our contributions include adapting AnoGAN's principles to a new domain and promising advancements in detecting previously undetectable anomalies. This paper delves into the multifaceted nature of anomaly detection, considering the dynamic evolution of normal behavior, context-dependent anomaly definitions, and data-related challenges like noise and imbalances.

Being able to successfully determine whether the testing samples has similar distribution as the training samples is a fundamental question to address before we can safely deploy most of the machine learning models into practice. In this paper, we propose TOOD detection, a simple yet effective tree-based out-of-distribution (TOOD) detection mechanism to determine if a set of unseen samples will have similar distribution as of the training samples. The TOOD detection mechanism is based on computing pairwise hamming distance of testing samples' tree embeddings, which are obtained by fitting a tree-based ensemble model through in-distribution training samples. Our approach is interpretable and robust for its tree-based nature. Furthermore, our approach is efficient, flexible to various machine learning tasks, and can be easily generalized to unsupervised setting. Extensive experiments are conducted to show the proposed method outperforms other state-of-the-art out-of-distribution detection methods in distinguishing the in-distribution from out-of-distribution on various tabular, image, and text data.

LLMs have been found to memorize training textual sequences and regurgitate verbatim said sequences during text generation time. This fact is known to be the cause of privacy and related (e.g., copyright) problems. Unlearning in LLMs then takes the form of devising new algorithms that will properly deal with these side-effects of memorized data, while not hurting the model's utility. We offer a fresh perspective towards this goal, namely, that each textual sequence to be forgotten should be treated differently when being unlearned based on its degree of memorization within the LLM. We contribute a new metric for measuring unlearning quality, an adversarial attack showing that SOTA algorithms lacking this perspective fail for privacy, and two new unlearning methods based on Gradient Ascent and Task Arithmetic, respectively. A comprehensive performance evaluation across an extensive suite of NLP tasks then mapped the solution space, identifying the best solutions under different scales in model capacities and forget set sizes and quantified the gains of the new approaches.

This systematic review focuses on anomaly detection for connected and autonomous vehicles. The initial database search identified 2160 articles, of which 203 were included in this review after rigorous screening and assessment. This study revealed that the most commonly used Artificial Intelligence (AI) algorithms employed in anomaly detection are neural networks like LSTM, CNN, and autoencoders, alongside one-class SVM. Most anomaly-based models were trained using real-world operational vehicle data, although anomalies, such as attacks and faults, were often injected artificially into the datasets. These models were evaluated mostly using five key evaluation metrics: recall, accuracy, precision, F1-score, and false positive rate. The most frequently used selection of evaluation metrics used for anomaly detection models were accuracy, precision, recall, and F1-score. This systematic review presents several recommendations. First, there is a need to incorporate multiple evaluation metrics to provide a comprehensive assessment of the anomaly detection models. Second, only a small proportion of the studies have made their models open source, indicating a need to share models publicly to facilitate collaboration within the research community, and to validate and compare findings effectively. Third, there is a need for benchmarking datasets with predefined anomalies or cyberattacks to test and improve the effectiveness of the proposed anomaly-based detection models. Furthermore, there is a need for future research to investigate the deployment of anomaly detection to a vehicle to assess its performance on the road. There is a notable lack of research done on intrusion detection systems using different protocols to CAN, such as Ethernet and FlexRay.