论文检索

Yang Gu, Jian Cao, Hengyu You, Nengjun Zhu, Shiyou Qian

The end-to-end automated design of machine learning (ML) pipelines significantly reduces the workload for data scientists and democratizes ML for non-experts. Evolutionary algorithm (EA)-based automated ML (AutoML) systems, a prominent category of AutoML, often face inefficiencies due to the costly fitness evaluation of candidate ML pipelines. Although surrogate models have been employed to approximate the true performance of pipelines more quickly, a key challenge remains in effectively bridging the semantic gap between the heterogeneous features of datasets and pipelines. To address this issue, we propose ADELA, a novel accompanying surrogate-based optimization strategy that accelerates EA-based AutoML while retaining the performance of the resulting pipelines. ADELA operates in two phases: Offline, leveraging a high-quality curated pipeline corpus to meta-learn an accompanying surrogate model; and Online, selecting the accompanying pipeline and using the learned model to predict the performance of evaluation pipelines instead of executing them. The accompanying mechanism effectively mitigates the semantic gap between datasets and pipelines, enabling ADELA to reduce computation times by an average of 73.66% while retaining 98.78% of the final pipeline performance, as demonstrated in extensive experimental evaluations.

Alessio Gravina, Moshe Eliasof, Claudio Gallicchio, Davide Bacciu, Carola-Bibiane Schönlieb

A common problem in Message-Passing Neural Networks is oversquashing -- the limited ability to facilitate effective information flow between distant nodes. Oversquashing is attributed to the exponential decay in information transmission as node distances increase. This paper introduces a novel perspective to address oversquashing, leveraging dynamical systems properties of global and local non-dissipativity, that enable the maintenance of a constant information flow rate. We present SWAN, a uniquely parameterized GNN model with antisymmetry both in space and weight domains, as a means to obtain non-dissipativity. Our theoretical analysis asserts that by implementing these properties, SWAN offers an enhanced ability to transmit information over extended distances. Empirical evaluations on synthetic and real-world benchmarks that emphasize long-range interactions validate the theoretical understanding of SWAN, and its ability to mitigate oversquashing.

Tony Gracious, Ambedkar Dukkipati

Understanding relations arising out of interactions among entities can be very difficult, and predicting them is even more challenging. This problem has many applications in various fields, such as financial networks and e-commerce. These relations can involve much more complexities than just involving more than two entities. One such scenario is evolving recursive relations between multiple entities, and so far, this is still an open problem. This work addresses the problem of forecasting higher-order interaction events that can be multi-relational and recursive. We pose the problem in the framework of representation learning of temporal hypergraphs that can capture complex relationships involving multiple entities. The proposed model, \textit{Relational Recursive Hyperedge Temporal Point Process} (RRHyperTPP) uses an encoder that learns a dynamic node representation based on the historical interaction patterns and then a hyperedge link prediction-based decoder to model the occurrence of interaction events. These learned representations are then used for downstream tasks involving forecasting the type and time of interactions. The main challenge in learning from hyperedge events is that the number of possible hyperedges grows exponentially with the number of nodes in the network. This will make the computation of negative log-likelihood of the temporal point process expensive, as the calculation of survival function requires a summation over all possible hyperedges. In our work, we develop a noise contrastive estimation method to learn the parameters of our model, and we have experimentally shown that our models perform better than previous state-of-the-art methods for interaction forecasting.

Tony Gracious, Arman Gupta, Ambedkar Dukkipati

Forecasting relations between entities is paramount in the current era of data and AI. However, it is often overlooked that real-world relationships are inherently directional, involve more than two entities, and can change with time. In this paper, we provide a comprehensive solution to the problem of forecasting directional relations in a general setting, where relations are higher-order, i.e., directed hyperedges in a hypergraph. This problem has not been previously explored in the existing literature. The primary challenge in solving this problem is that the number of possible hyperedges is exponential in the number of nodes at each event time. To overcome this, we propose a sequential generative approach that segments the forecasting process into multiple stages, each contingent upon the preceding stages, thereby reducing the search space involved in predictions of hyperedges. The first stage involves a temporal point process-based node event forecasting module that identifies the subset of nodes involved in an event. The second stage is a candidate generation module that predicts hyperedge sizes and adjacency vectors for nodes observing events. The final stage is a directed hyperedge predictor that identifies the truth by searching over the set of candidate hyperedges. To validate the effectiveness of our model, we compiled five datasets and conducted an extensive empirical study to assess each downstream task. Our proposed method achieves a performance gain of 32% and 41% compared to the state-of-the-art pairwise and hyperedge event forecasting models, respectively, for the event type prediction.

Ze Gong, Akshat Kumar, Pradeep Varakantham

Offline safe reinforcement learning (RL) has emerged as a promising approach for learning safe behaviors without engaging in risky online interactions with the environment. Most existing methods in offline safe RL rely on cost constraints at each time step (derived from global cost constraints) and this can result in either overly conservative policies or violation of safety constraints. In this paper, we propose to learn a policy that generates desirable trajectories and avoids undesirable trajectories. To be specific, we first partition the pre-collected dataset of state-action trajectories into desirable and undesirable subsets. Intuitively, the desirable set contains high reward and safe trajectories, and undesirable set contains unsafe trajectories and low-reward safe trajectories. Second, we learn a policy that generates desirable trajectories and avoids undesirable trajectories, where (un)desirability scores are provided by a classifier learnt from the dataset of desirable and undesirable trajectories. This approach bypasses the computational complexity and stability issues of a min-max objective that is employed in existing methods. Theoretically, we also show our approach’s strong connections to existing learning paradigms involving human feedback. Finally, we extensively evaluate our method using the DSRL benchmark for offline safe RL. Empirically, our method outperforms competitive baselines, achieving higher rewards and better constraint satisfaction across a wide variety of benchmark tasks.

Xueping Gong, Jiheng Zhang

This paper presents a novel framework for multi-armed bandit problems with side-observations and switching constraints, which arises in a range of real-world applications such as robotic. To address the challenges of effectively utilizing graph-structured observations while adhering to graph constraints, we design graph-agnostic and graph-aware algorithms tailored to this new setting. Specifically, our graph-agnostic algorithm selects nodes with the highest upper confidence bound without prior knowledge of feedback probabilities, while minimizing switching costs using offline shortest path planning and the doubling trick. If the graph structure and associated probability matrix are known, our graph-aware algorithm plans the exploration step using a linear programming approach and eliminates suboptimal nodes iteratively. We rigorously analyze the performance of our proposed algorithms, providing near-optimal minimax and instance-dependent regret upper bounds. Our analysis shows that our algorithms outperform generic reinforcement learning methods in terms of both regret and computational efficiency. Extensive numerical experiments on various types of graphs, including two real-world datasets, demonstrate the efficacy of our proposed methods and their advantages over benchmark methods in graph bandit settings.

Xiuwen Gong, Nitin Bisht, Guandong Xu

Partial label learning (PLL) allows each instance to be annotated with a set of candidate labels, but only one is the ground-truth label. Although the state-of-the-art (SOTA) PLL models have shown competitive performance, they cannot get rid of the negative influence from the noisy false-positive labels during the training process. This leads to a large extent of uncertainty of PLL models’ prediction, and it becomes unreliable to trust a PLL model’s performance only by its prediction accuracy. To bridge this gap, we develop a new framework to quantify the uncertainty for PLL models with valid confidence guarantee, which is named as Conformal Prediction for Partial Label Learning (CP-PLL). This framework can be implemented on top of any PLL method to quantify their predictive confidence in terms of average prediction set size with a use-specified error rate or coverage/confidence level (i.e., probability). We prove that the coverage guarantee in PLL still holds, that is, the ground-truth label can be covered in the constructed prediction set with the user pre-defined error rate α when we use the noisy calibration data to carlibrate the PLL models, which yields to a probability interval of [1- α, 1- α + 1/n+1 + ε]. Extensive experiments are conducted on SOTA PLL methods and benchmark datasets to verify the effectiveness of the proposed framework.

Wenwen Gong, Yangliao Geng, Dan Zhang, Yifan Zhu, Xiaolong Xu, Haolong Xiang, Amin Beheshti, Xuyun Zhang, Lianyong Qi

Graph Contrastive Learning (GCL), as a primary paradigm of graph self-supervised learning, spurs a fruitful line of research in tackling the data sparsity issue by maximizing the consistency of user/item embeddings between different augmented views with random perturbations. However, diversity, as a crucial metric for recommendation performance and user satisfaction, has received rather little attention. In fact, there exists a challenging dilemma in balancing accuracy and diversity. To address these issues, we propose a new Graph Contrastive Learning (DivGCL) model for diversifying recommendations. Inspired by the excellence of the determinant point process (DPP), DivGCL adopts a DPP likelihood-based loss function to achieve an ideal trade-off between diversity and accuracy, optimizing it jointly with the advanced Gaussian noise-augmented GCL objective. Extensive experiments on four popular datasets demonstrate that DivGCL surpasses existing approaches in balancing accuracy and diversity, with an improvement of 23.47% at T@20 (abbreviation for trade-off metric) on ML-1M.

Nanxu Gong, Chandan K Reddy, Wangyang Ying, Haifeng Chen, Yanjie Fu

Feature transformation aims to reconstruct the feature space of raw features to enhance the performance of downstream models. However, the exponential growth in the combinations of features and operations poses a challenge, making it difficult for existing methods to efficiently explore a wide space. Additionally, their optimization is solely driven by the accuracy of downstream models in specific domains, neglecting the acquisition of general feature knowledge. To fill this research gap, we propose an evolutionary LLM framework for automated feature transformation. This framework consists of two parts: 1) constructing a multi-population database through an RL data collector while utilizing evolutionary algorithm strategies for database maintenance, and 2) utilizing the ability of Large Language Model (LLM) in sequence understanding, we employ few-shot prompts to guide LLM in generating superior samples based on feature transformation sequence distinction. Leveraging the multi-population database initially provides a wide search scope to discover excellent populations. Through culling and evolution, high-quality populations are given greater opportunities, thereby furthering the pursuit of optimal individuals. By integrating LLMs with evolutionary algorithms, we achieve efficient exploration within a vast space, while harnessing feature knowledge to propel optimization, thus realizing a more adaptable search paradigm. Finally, we empirically demonstrate the effectiveness and generality of our proposed method.

Mingrong Gong, Chaoqi Chen, Qingqiang Sun, Yue Wang, Hui Huang

Out-of-distribution (OOD) detection is a crucial task for deploying deep learning models in the wild. One of the major challenges is that well-trained deep models tend to perform over-confidence on unseen test data. Recent research attempts to leverage real or synthetic outliers to mitigate the issue, which may significantly increase computational costs and be biased toward specific outlier characteristics. In this paper, we propose a simple yet effective framework, Prototypical Outlier Proxy (POP), which introduces virtual OOD prototypes to reshape the decision boundaries between ID and OOD data. Specifically, we transform the learnable classifier into a fixed one and augment it with a set of prototypical weight vectors. Then, we introduce a hierarchical similarity boundary loss to impose adaptive penalties depending on the degree of misclassification. Extensive experiments across various benchmarks demonstrate the effectiveness of POP. Notably, POP achieves average FPR95 reductions of 7.70%, 6.30%, and 5.42% over the second-best methods on CIFAR-10, CIFAR-100, and ImageNet-200, respectively. Moreover, compared to the recent method NPOS, which relies on outlier synthesis, POP trains 7.2 times faster and performs inference 19.5 times faster.

Fengjiiao Gong, Angxiao Yue, Hongteng Xu

Multi-modal learning aims to learn predictive models based on the data from different modalities. However, due to the requirement of data security and privacy protection, real-world multi-modal data are often scattered to different agents and cannot be shared across the agents, which limits the application of existing multi-modal learning methods. To achieve robust multi-modal learning in such a challenging scenario, we propose a novel optimal transport-based mixer (OTM), which works as an effective latent code alignment and augmentation method for unaligned and distributed multi-modal data. In particular, we train a Wasserstein autoencoder (WAE) for each agent, which encodes its single modal samples in a latent space. Through a central server, the proposed OTM computes a stochastic fused Gromov-Wasserstein barycenter (FGWB) to mix different modalities' latent codes, so that each agent applies the barycenter to reconstruct its samples. This method neither requires well-aligned multi-modal data nor assumes the data to share the same latent distribution, and each agent can learn a specific model based on multi-modal data while achieving inference based on its local modality. Experiments on multi-modal clustering and classification demonstrate that the models learned with the OTM method outperform the corresponding baselines.

Augustin Godinot, Erwan Le Merrer, Camilla Penzo, Francois Taiani, Gilles Tredan

The deployment of machine learning models in operational contexts represents a significant investment for any organisation. Consequently, the risk of these models being misappropriated by competitors needs to be addressed. In recent years, numerous proposals have been put forth to detect instances of model stealing. However, these proposals operate under implicit and disparate data and model access assumptions; as a consequence, it remains unclear how they can be effectively compared to one another. Our evaluation shows that a simple baseline that we introduce performs on par with existing state-of-the-art fingerprints, which, on the other hand, are much more complex. To uncover the reasons behind this intriguing result, this paper introduces a systematic approach to both the creation of model fingerprinting schemes and their evaluation benchmarks. By dividing model fingerprinting into three core components – Query, Representation and Detection (QuRD) – we are able to identify ~100 previously unexplored QuRD combinations and gain insights into their performance. Finally, we introduce a set of metrics to compare and guide the creation of more representative model stealing detection benchmarks. Our approach reveals the need for more challenging benchmarks and a sound comparison with baselines. To foster the creation of new fingerprinting schemes and benchmarks, we open-source our fingerprinting toolbox.

Mohsen Ghassemi, Alan Mishler, Niccolo Dalmasso, Luhao Zhang, Vamsi K. Potluru, Tucker Balch, Manuela Veloso

Conditional demographic parity (CDP) is a measure of the demographic parity of a predictive model or decision process when conditioning on an additional feature or set of features. Many algorithmic fairness techniques exist to target demographic parity, but CDP is much harder to achieve, particularly when the conditioning variable has many levels and/or when the model outputs are continuous. The problem of auditing and enforcing CDP is understudied in the literature. In light of this, we propose novel measures of conditional demographic disparity (CDD) which rely on statistical distances borrowed from the optimal transport literature. We further design and evaluate regularization-based approaches based on these CDD measures. Our methods, FairBiT and FairLeap, allow us to target conditional demographic parity even when the conditioning variable has many levels. When model outputs are continuous, our methods target full equality of the conditional distributions, unlike other methods that only consider first moments or related proxy quantities. We validate our approaches on real-world datasets.

Paul Ghanem, Ahmet Demirkaya, Tales Imbiriba, Alireza Ramezani, Zachary Danziger, Deniz Erdogmus

Learning dynamics governing physical and spatiotemporal processes is a challenging problem, especially in scenarios where states are partially measured. In this work, we tackle the problem of learning dynamics governing these systems when parts of the system's states are not measured, specifically when the dynamics generating the non-measured states are unknown. Inspired by state estimation theory and Physics Informed Neural ODEs, we present a sequential optimization framework in which dynamics governing unmeasured processes can be learned. We demonstrate the performance of the proposed approach leveraging both numerical simulations and a real dataset extracted from an electro-mechanical positioning system. We show how the underlying equations fit into our formalism and demonstrate the improved performance of the proposed method when compared with standard baselines.

Xiwen Geng, Suyun Zhao, Yixin Yu, Borui Peng, Pan Du, Hong Chen, Cuiping Li, Mengdie Wang

Clustering traditionally aims to reveal a natural grouping structure within unlabeled data. However, this structure may not always align with users' preferences. In this paper, we propose a personalized clustering method that explicitly performs targeted representation learning by interacting with users via modicum task information (e.g., must-link or cannot-link pairs) to guide the clustering direction. We query users with the most informative pairs, i.e., those pairs most hard to cluster and those most easy to miscluster, to facilitate the representation learning in terms of the clustering preference. Moreover, by exploiting attention mechanism, the targeted representation is learned and augmented. By leveraging the targeted representation and constrained contrastive loss as well, personalized clustering is obtained. Theoretically, we verify that the risk of personalized clustering is tightly bounded, guaranteeing that active queries to users do mitigate the clustering risk. Experimentally, extensive results show that our method performs well across different clustering tasks and datasets, even when only a limited number of queries are available.

Eloy Geenjaar, Lie Lu

Transfer learning for bio-signals has recently become an important technique to improve prediction performance on downstream tasks with small bio-signal datasets. Recent works have shown that pre-training a neural network model on a large dataset (e.g. EEG) with a self-supervised task, replacing the self-supervised head with a linear classification head, and fine-tuning the model on different downstream bio-signal datasets (e.g., EMG or ECG) can dramatically improve the performance on those datasets. In this paper, we propose a new convolution-transformer hybrid model architecture with masked auto-encoding for low-data bio-signal transfer learning, introduce a frequency-based masked auto-encoding task, employ a more comprehensive evaluation framework, and evaluate how much and when (multimodal) pre-training improves fine-tuning performance. We also introduce a dramatically more performant method of aligning a downstream dataset with a different temporal length and sampling rate to the original pre-training dataset. Our findings indicate that the convolution-only part of our hybrid model can achieve state-of-the-art performance on some low-data downstream tasks. The performance is often improved even further with our full model. In the case of transformer-based models we find that pre-training especially improves performance on downstream datasets, multimodal pre-training often increases those gains further, and our frequency-based pre-training performs the best on average for the lowest and highest data regimes.

Jorge García-Carrasco, Alejandro Maté, Juan Trujillo

Large Language Models (LLMs) have shown impressive performance across a wide range of tasks. However, the size of LLMs is steadily increasing, hindering their application on computationally constrained environments. On the other hand, despite their general capabilities, there are many situations where only one specific task is performed, rendering all other capabilities unnecessary and wasteful. This leads us to the following question: Is it possible to extract the minimal subset from an LLM that is able to perform a specific task in a faster, standalone manner? Recent works on Mechanistic Interpretability (MI) have shown that specific tasks are performed by a localized subset of components, or circuit. However, current techniques used to identify the circuit cannot be used to extract it for its standalone usage. In this work, we propose a novel approach to automatically extract the subset of the LLM that properly performs a targeted task requiring no additional training and a small amount of data samples. We evaluate our approach on different tasks and show that the resulting models are (i) considerably smaller, reducing the number of parameters up to 82.77% and (ii) more interpretable, as they focus on the circuit that is used to carry out the specific task, and can therefore be understood using MI techniques.

Zijian Gao, Shanhao Han, Xingxing Zhang, Kele Xu, Dulan Zhou, Xinjun Mao, Yong Dou, Huaimin Wang

Logit-based knowledge distillation (KD) is commonly used to mitigate catastrophic forgetting in class-incremental learning (CIL) caused by data distribution shifts. However, the strict match of logit values between student and teacher models conflicts with the cross-entropy (CE) loss objective of learning new classes, leading to significant recency bias (i.e. unfairness). To address this issue, we rethink the overlooked limitations of KD-based methods through empirical analysis. Inspired by our findings, we introduce a plug-and-play pre-process method that normalizes the logits of both the student and teacher across all classes, rather than just the old classes, before distillation. This approach allows the student to focus on both old and new classes, capturing intrinsic inter-class relations from the teacher. By doing so, our method avoids the inherent conflict between KD and CE, maintaining fairness between old and new classes. Additionally, recognizing that overconfident teacher predictions can hinder the transfer of inter-class relations (i.e., dark knowledge), we extend our method to capture intra-class relations among different instances, ensuring fairness within old classes. Our method integrates seamlessly with existing logit-based KD approaches, consistently enhancing their performance across multiple CIL benchmarks without incurring additional training costs.

Xiyuan Gao, Bing Cao, Pengfei Zhu, Nannan Wang, Qinghua Hu

The strength of multimodal learning lies in its ability to integrate information from various sources, providing rich and comprehensive insights. However, in real-world scenarios, multi-modal systems often face the challenge of dynamic modality contributions, the dominance of different modalities may change with the environments, leading to suboptimal performance in multimodal learning. Current methods mainly enhance weak modalities to balance multimodal representation bias, which inevitably optimizes from a partialmodality perspective, easily leading to performance descending for dominant modalities. To address this problem, we propose an Asymmetric Reinforcing method against Multimodal representation bias (ARM). Our ARM dynamically reinforces the weak modalities while maintaining the ability to represent dominant modalities through conditional mutual information. Moreover, we provide an in-depth analysis that optimizing certain modalities could cause information loss and prevent leveraging the full advantages of multimodal data. By exploring the dominance and narrowing the contribution gaps between modalities, we have significantly improved the performance of multimodal learning, making notable progress in mitigating imbalanced multimodal learning.

Shiqi Gao, Tianxiang Gong, Zijie Lin, Runhua Xu, Haoyi Zhou, Jianxin Li

Uncertainty estimation is essential for practical applications such as decision-making, risk assessment, and human-AI collaboration. However, Uncertainty estimation in open-ended question-answering (QA) tasks presents unique challenges. The output space for open-ended QA is vast and discrete, and the autoregressive nature of LLMs, combined with the rapid increase in model parameters, makes inference sampling significantly costly. An ideal uncertainty estimation for LLMs should meet two criteria: 1) incur no additional inference cost and 2) capture the semantic dependencies of token-level uncertainty within sequences. We propose a promising solution that converts redundancy into randomness in the extensive parameters of LLMs to quantify knowledge uncertainty. We can obtain token-level Monte Carlo samples without multiple inferences by introducing randomness during a single forward pass. We theoretically analyze the FLUE sampling method and employ a post-processing method to learn the state transitions from token uncertainty to sequence uncertainty. In open-ended question-answering tasks, we demonstrate that FLUE can achieve competitive performance in estimating the uncertainty of generated sentences without adding extra inference overhead.