While personalized recommendations systems have become increasingly popular, ensuring user data protection remains a top concern in the development of these learning systems. A common approach to enhancing privacy involves training models using anonymous data rather than individual data. In this paper, we explore a natural technique called look-alike clustering, which involves replacing sensitive features of individuals with the cluster's average values. We provide a precise analysis of how training models using anonymous cluster centers affects their generalization capabilities. We focus on an asymptotic regime where the size of the training set grows in proportion to the features dimension. Our analysis is based on the Convex Gaussian Minimax Theorem (CGMT) and allows us to theoretically understand the role of different model components on the generalization error. In addition, we demonstrate that in certain high-dimensional regimes, training over anonymous cluster centers acts as a regularization and improves generalization error of the trained models. Finally, we corroborate our asymptotic theory with finite-sample numerical experiments where we observe a perfect match when the sample size is only of order of a few hundreds.

Distributionally robust learning (DRL) is increasingly seen as a viable method to train machine learning models for improved model generalization. These min-max formulations, however, are more difficult to solve. We provide a new stochastic gradient descent algorithm to efficiently solve this DRL formulation. Our approach applies gradient descent to the outer minimization formulation and estimates the gradient of the inner maximization based on a sample average approximation. The latter uses a subset of the data sampled without replacement in each iteration, progressively increasing the subset size to ensure convergence. We rigorously establish convergence to a near-optimal solution under standard regularity assumptions and, for strongly convex losses, match the best known $O(\epsilon{ 1})$ rate of convergence up to a known threshold. Empirical results demonstrate the significant benefits of our approach over previous work in improving learning for model generalization.

Data augmentation plays a crucial role in enhancing model performance across various AI fields by introducing variability while maintaining the underlying temporal patterns. However, in the context of long sequence time series data, where maintaining temporal consistency is critical, there are fewer augmentation strategies compared to fields such as image or text, with advanced techniques like Mixup rarely being used. In this work, we propose a new approach, Imputation-based Mixup Augmentation (IMA), which combines Imputed-data Augmentation with Mixup Augmentation to bolster model generalization and improve forecasting performance. We evaluate the effectiveness of this method across several forecasting models, including DLinear (MLP), TimesNet (CNN), and iTrainformer (Transformer), these models represent some of the most recent advances in long sequence time series forecasting. Our experiments, conducted on three datasets (ETT -small, Illness, Exchange Rate) from various domains and compared against eight other augmentation techniques, demonstrate that IMA consistently enhances performance, achieving 22 improvements out of 24 instances, with 10 of those being the best performances, particularly with iTrain-former imputation in ETT dataset. The GitHub repository is available at: https://github.com/dangnha/IMA.

Deep learned (DL) models for image recognition have been shown to fail to generalize to data from different devices, populations, etc. COVID-19 detection from Chest X-rays (CXRs), in particular, has been shown to fail to generalize to out-of-distribution (OOD) data from new clinical sources not covered in the training set. This occurs because models learn to exploit shortcuts - source-specific artifacts that do not translate to new distributions - rather than reasonable biomarkers to maximize performance on in-distribution (ID) data. Rendering the models more robust to distribution shifts, our study investigates the use of fundamental noise injection techniques (Gaussian, Speckle, Poisson, and Salt and Pepper) during training. Our empirical results demonstrate that this technique can significantly reduce the performance gap between ID and OOD evaluation from 0.10 0.20 to 0.01 0.06, based on results averaged over ten random seeds across key metrics such as AUC, F1, accuracy, recall and specificity.

The growing popularity of large language models has raised concerns regarding the potential to misuse AI-generated text (AIGT). It becomes increasingly critical to establish an excellent AIGT detection method with high generalization and robustness. However, existing methods either focus on model generalization or concentrate on robustness. The unified mechanism, to simultaneously address the challenges of generalization and robustness, is less explored. In this paper, we argue that robustness can be view as a specific form of domain shift, and empirically reveal an intrinsic mechanism for model generalization of AIGT detection task. Then, we proposed a novel AIGT detection method (DP-Net) via dynamic perturbations introduced by a reinforcement learning with elaborated reward and action. Experimentally, extensive results show that the proposed DP-Net significantly outperforms some state-of-the-art AIGT detection methods for generalization capacity in three cross-domain scenarios. Meanwhile, the DP-Net achieves best robustness under two text adversarial attacks. The code is publicly available at https://github.com/CAU-ISS-Lab/AIGT-Detection-Evade-Detection/tree/main/DP-Net.

Existing saliency-guided training approaches improve model generalization by incorporating a loss term that compares the model's class activation map (CAM) for a sample's true-class ({\it i.e.}, correct-label class) against a human reference saliency map. However, prior work has ignored the false-class CAM(s), that is the model's saliency obtained for incorrect-label class. We hypothesize that in binary tasks the true and false CAMs should diverge on the important classification features identified by humans (and reflected in human saliency maps). We use this hypothesis to motivate three new saliency-guided training methods incorporating both true- and false-class model's CAM into the training strategy and a novel post-hoc tool for identifying important features. We evaluate all introduced methods on several diverse binary close-set and open-set classification tasks, including synthetic face detection, biometric presentation attack detection, and classification of anomalies in chest X-ray scans, and find that the proposed methods improve generalization capabilities of deep learning models over traditional (true-class CAM only) saliency-guided training approaches. We offer source codes and model weights\footnote{GitHub repository link removed to preserve anonymity} to support reproducible research.

Human computer interaction has become integral to modern life, driven by advancements in machine learning technologies. Affective computing, in particular, has focused on systems that recognize, interpret, and respond to human emotions, often using wearable devices, which provide continuous data streams of physiological signals. Among various physiological signals, the photoplethysmogram (PPG) has gained prominence due to its ease of acquisition from widely available devices. However, the generalization of PPG-based emotion recognition models across individuals remains an unresolved challenge. This paper introduces a novel hybrid architecture that combines Convolutional Neural Networks (CNNs), Long Short-Term Memory networks (LSTMs), and Temporal Convolutional Networks (TCNs) to address this issue. The proposed model integrates the strengths of these architectures to improve robustness and generalization. Raw PPG signals are fed into the CNN for feature extraction. These features are processed separately by LSTM and TCN. The outputs from these components are concatenated to generate a final feature representation, which serves as the input for classifying valence and arousal, the primary dimensions of emotion. Experiments using the Photoplethysmogram Dataset for Emotional Analysis (PPGE) demonstrate that the proposed hybrid model achieves better model generalization than standalone CNN and LSTM architectures. Our results show that the proposed solution outperforms the state-of-the-art CNN architecture, as well as a CNN-LSTM model, in emotion recognition tasks with PPG signals. Using metrics such as Area Under the Curve (AUC) and F1 Score, we highlight the model's effectiveness in handling subject variability.