Machine Learning (cs.LG)

Abstract: Chronic stress can significantly affect physical and mental health. The advent of wearable technology allows for the tracking of physiological signals, potentially leading to innovative stress prediction and intervention methods. However, challenges such as label scarcity and data heterogeneity render stress prediction difficult in practice. To counter these issues, we have developed a multimodal personalized stress prediction system using wearable biosignal data. We employ self-supervised learning (SSL) to pre-train the models on each subject's data, allowing the models to learn the baseline dynamics of the participant's biosignals prior to fine-tuning the stress prediction task. We test our model on the Wearable Stress and Affect Detection (WESAD) dataset, demonstrating that our SSL models outperform non-SSL models while utilizing less than 5% of the annotations. These results suggest that our approach can personalize stress prediction to each user with minimal annotations. This paradigm has the potential to enable personalized prediction of a variety of recurring health events using complex multimodal data streams.

Abstract: For many real-world time series tasks, the computational complexity of prevalent deep leaning models often hinders the deployment on resource-limited environments (e.g., smartphones). Moreover, due to the inevitable domain shift between model training (source) and deploying (target) stages, compressing those deep models under cross-domain scenarios becomes more challenging. Although some of existing works have already explored cross-domain knowledge distillation for model compression, they are either biased to source data or heavily tangled between source and target data. To this end, we design a novel end-to-end framework called Universal and joint knowledge distillation (UNI-KD) for cross-domain model compression. In particular, we propose to transfer both the universal feature-level knowledge across source and target domains and the joint logit-level knowledge shared by both domains from the teacher to the student model via an adversarial learning scheme. More specifically, a feature-domain discriminator is employed to align teacher's and student's representations for universal knowledge transfer. A data-domain discriminator is utilized to prioritize the domain-shared samples for joint knowledge transfer. Extensive experimental results on four time series datasets demonstrate the superiority of our proposed method over state-of-the-art (SOTA) benchmarks.

Abstract: User selection has become crucial for decreasing the communication costs of federated learning (FL) over wireless networks. However, centralized user selection causes additional system complexity. This study proposes a network intrinsic approach of distributed user selection that leverages the radio resource competition mechanism in random access. Taking the carrier sensing multiple access (CSMA) mechanism as an example of random access, we manipulate the contention window (CW) size to prioritize certain users for obtaining radio resources in each round of training. Training data bias is used as a target scenario for FL with user selection. Prioritization is based on the distance between the newly trained local model and the global model of the previous round. To avoid excessive contribution by certain users, a counting mechanism is used to ensure fairness. Simulations with various datasets demonstrate that this method can rapidly achieve convergence similar to that of the centralized user selection approach.

Abstract: Many machine learning tasks can be formulated as a stochastic compositional optimization (SCO) problem such as reinforcement learning, AUC maximization, and meta-learning, where the objective function involves a nested composition associated with an expectation. While a significant amount of studies has been devoted to studying the convergence behavior of SCO algorithms, there is little work on understanding their generalization, i.e., how these learning algorithms built from training examples would behave on future test examples. In this paper, we provide the stability and generalization analysis of stochastic compositional gradient descent algorithms through the lens of algorithmic stability in the framework of statistical learning theory. Firstly, we introduce a stability concept called compositional uniform stability and establish its quantitative relation with generalization for SCO problems. Then, we establish the compositional uniform stability results for two popular stochastic compositional gradient descent algorithms, namely SCGD and SCSC. Finally, we derive dimension-independent excess risk bounds for SCGD and SCSC by trade-offing their stability results and optimization errors. To the best of our knowledge, these are the first-ever-known results on stability and generalization analysis of stochastic compositional gradient descent algorithms.

Abstract: Anomaly detection based on system logs plays an important role in intelligent operations, which is a challenging task due to the extremely complex log patterns. Existing methods detect anomalies by capturing the sequential dependencies in log sequences, which ignore the interactions of subsequences. To this end, we propose CSCLog, a Component Subsequence Correlation-Aware Log anomaly detection method, which not only captures the sequential dependencies in subsequences, but also models the implicit correlations of subsequences. Specifically, subsequences are extracted from log sequences based on components and the sequential dependencies in subsequences are captured by Long Short-Term Memory Networks (LSTMs). An implicit correlation encoder is introduced to model the implicit correlations of subsequences adaptively. In addition, Graph Convolution Networks (GCNs) are employed to accomplish the information interactions of subsequences. Finally, attention mechanisms are exploited to fuse the embeddings of all subsequences. Extensive experiments on four publicly available log datasets demonstrate the effectiveness of CSCLog, outperforming the best baseline by an average of 7.41% in Macro F1-Measure.

Abstract: The increasing concerns regarding the privacy of machine learning models have catalyzed the exploration of machine unlearning, i.e., a process that removes the influence of training data on machine learning models. This concern also arises in the realm of federated learning, prompting researchers to address the federated unlearning problem. However, federated unlearning remains challenging. Existing unlearning methods can be broadly categorized into two approaches, i.e., exact unlearning and approximate unlearning. Firstly, implementing exact unlearning, which typically relies on the partition-aggregation framework, in a distributed manner does not improve time efficiency theoretically. Secondly, existing federated (approximate) unlearning methods suffer from imprecise data influence estimation, significant computational burden, or both. To this end, we propose a novel federated unlearning framework based on incremental learning, which is independent of specific models and federated settings. Our framework differs from existing federated unlearning methods that rely on approximate retraining or data influence estimation. Instead, we leverage new memories to overwrite old ones, imitating the process of \textit{active forgetting} in neurology. Specifically, the model, intended to unlearn, serves as a student model that continuously learns from randomly initiated teacher models. To preserve catastrophic forgetting of non-target data, we utilize elastic weight consolidation to elastically constrain weight change. Extensive experiments on three benchmark datasets demonstrate the efficiency and effectiveness of our proposed method. The result of backdoor attacks demonstrates that our proposed method achieves satisfying completeness.

Abstract: This work introduces a novel approach to pruning deep learning models by using distilled data. Unlike conventional strategies which primarily focus on architectural or algorithmic optimization, our method reconsiders the role of data in these scenarios. Distilled datasets capture essential patterns from larger datasets, and we demonstrate how to leverage this capability to enable a computationally efficient pruning process. Our approach can find sparse, trainable subnetworks (a.k.a. Lottery Tickets) up to 5x faster than Iterative Magnitude Pruning at comparable sparsity on CIFAR-10. The experimental results highlight the potential of using distilled data for resource-efficient neural network pruning, model compression, and neural architecture search.

Abstract: Multi-Task Learning (MTL) is a powerful technique that has gained popularity due to its performance improvement over traditional Single-Task Learning (STL). However, MTL is often challenging because there is an exponential number of possible task groupings, which can make it difficult to choose the best one, and some groupings might produce performance degradation due to negative interference between tasks. Furthermore, existing solutions are severely suffering from scalability issues, limiting any practical application. In our paper, we propose a new data-driven method that addresses these challenges and provides a scalable and modular solution for classification task grouping based on hand-crafted features, specifically Data Maps, which capture the training behavior for each classification task during the MTL training. We experiment with the method demonstrating its effectiveness, even on an unprecedented number of tasks (up to 100).

Abstract: Large language models like GPT-4 exhibit emergent capabilities across general-purpose tasks, such as basic arithmetic, when trained on extensive text data, even though these tasks are not explicitly encoded by the unsupervised, next-token prediction objective. This study investigates how small transformers, trained from random initialization, can efficiently learn arithmetic operations such as addition, multiplication, and elementary functions like square root, using the next-token prediction objective. We first demonstrate that conventional training data is not the most effective for arithmetic learning, and simple formatting changes can significantly improve accuracy. This leads to sharp phase transitions as a function of training data scale, which, in some cases, can be explained through connections to low-rank matrix completion. Building on prior work, we then train on chain-of-thought style data that includes intermediate step results. Even in the complete absence of pretraining, this approach significantly and simultaneously improves accuracy, sample complexity, and convergence speed. We also study the interplay between arithmetic and text data during training and examine the effects of few-shot prompting, pretraining, and model scale. Additionally, we discuss length generalization challenges. Our work highlights the importance of high-quality, instructive data that considers the particular characteristics of the next-word prediction objective for rapidly eliciting arithmetic capabilities.