Machine Learning (cs.LG)

Abstract: With the emerging developments of the Metaverse, a virtual world where people can interact, socialize, play, and conduct their business, it has become critical to ensure that the underlying systems are transparent, secure, and trustworthy. To this end, we develop a decentralized and trustworthy quantum federated learning (QFL) framework. The proposed QFL leverages the power of blockchain to create a secure and transparent system that is robust against cyberattacks and fraud. In addition, the decentralized QFL system addresses the risks associated with a centralized server-based approach. With extensive experiments and analysis, we evaluate classical federated learning (CFL) and QFL in a distributed setting and demonstrate the practicality and benefits of the proposed design. Our theoretical analysis and discussions develop a genuinely decentralized financial system essential for the Metaverse. Furthermore, we present the application of blockchain-based QFL in a hybrid metaverse powered by a metaverse observer and world model. Our implementation details and code are publicly available 1.

Abstract: We study a search and tracking (S&T) problem for a team of dynamic search agents to capture an adversarial evasive agent with only sparse temporal and spatial knowledge of its location in this paper. The domain is challenging for traditional Reinforcement Learning (RL) approaches as the large space leads to sparse observations of the adversary and in turn sparse rewards for the search agents. Additionally, the opponent's behavior is reactionary to the search agents, which causes a data distribution shift for RL during training as search agents improve their policies. We propose a differentiable Multi-Agent RL (MARL) architecture that utilizes a novel filtering module to supplement estimated adversary location information and enables the effective learning of a team policy. Our algorithm learns how to balance information from prior knowledge and a motion model to remain resilient to the data distribution shift and outperforms all baseline methods with a 46% increase of detection rate.

Abstract: Deep ensemble is a simple yet powerful way to improve the performance of deep neural networks. Under this motivation, recent works on mode connectivity have shown that parameters of ensembles are connected by low-loss subspaces, and one can efficiently collect ensemble parameters in those subspaces. While this provides a way to efficiently train ensembles, for inference, multiple forward passes should still be executed using all the ensemble parameters, which often becomes a serious bottleneck for real-world deployment. In this work, we propose a novel framework to reduce such costs. Given a low-loss subspace connecting two modes of a neural network, we build an additional neural network that predicts the output of the original neural network evaluated at a certain point in the low-loss subspace. The additional neural network, which we call a "bridge", is a lightweight network that takes minimal features from the original network and predicts outputs for the low-loss subspace without forward passes through the original network. We empirically demonstrate that we can indeed train such bridge networks and significantly reduce inference costs with the help of bridge networks.

Abstract: Graph Representation Learning (GRL) is an influential methodology, enabling a more profound understanding of graph-structured data and aiding graph clustering, a critical task across various domains. The recent incursion of attention mechanisms, originally an artifact of Natural Language Processing (NLP), into the realm of graph learning has spearheaded a notable shift in research trends. Consequently, Graph Attention Networks (GATs) and Graph Attention Auto-Encoders have emerged as preferred tools for graph clustering tasks. Yet, these methods primarily employ a local attention mechanism, thereby curbing their capacity to apprehend the intricate global dependencies between nodes within graphs. Addressing these impediments, this study introduces an innovative method known as the Graph Transformer Auto-Encoder for Graph Clustering (GTAGC). By melding the Graph Auto-Encoder with the Graph Transformer, GTAGC is adept at capturing global dependencies between nodes. This integration amplifies the graph representation and surmounts the constraints posed by the local attention mechanism. The architecture of GTAGC encompasses graph embedding, integration of the Graph Transformer within the autoencoder structure, and a clustering component. It strategically alternates between graph embedding and clustering, thereby tailoring the Graph Transformer for clustering tasks, whilst preserving the graph's global structural information. Through extensive experimentation on diverse benchmark datasets, GTAGC has exhibited superior performance against existing state-of-the-art graph clustering methodologies. This pioneering approach represents a novel contribution to the field of graph clustering, paving the way for promising avenues in future research.

Abstract: With the explosion of graph-structured data, link prediction has emerged as an increasingly important task. Embedding methods for link prediction utilize neural networks to generate node embeddings, which are subsequently employed to predict links between nodes. However, the existing embedding methods typically take a holistic strategy to learn node embeddings and ignore the entanglement of latent factors. As a result, entangled embeddings fail to effectively capture the underlying information and are vulnerable to irrelevant information, leading to unconvincing and uninterpretable link prediction results. To address these challenges, this paper proposes a novel framework with two variants, the disentangled graph auto-encoder (DGAE) and the variational disentangled graph auto-encoder (VDGAE). Our work provides a pioneering effort to apply the disentanglement strategy to link prediction. The proposed framework infers the latent factors that cause edges in the graph and disentangles the representation into multiple channels corresponding to unique latent factors, which contributes to improving the performance of link prediction. To further encourage the embeddings to capture mutually exclusive latent factors, we introduce mutual information regularization to enhance the independence among different channels. Extensive experiments on various real-world benchmarks demonstrate that our proposed methods achieve state-of-the-art results compared to a variety of strong baselines on link prediction tasks. Qualitative analysis on the synthetic dataset also illustrates that the proposed methods can capture distinct latent factors that cause links, providing empirical evidence that our models are able to explain the results of link prediction to some extent. All code will be made publicly available upon publication of the paper.

Abstract: To ensure resilient neural network processing on even unreliable hardware, comprehensive reliability analysis against various hardware faults is generally required before the deep neural network models are deployed, and efficient error injection tools are highly demanded. However, most existing fault injection tools remain rather limited to basic fault injection to neurons and fail to provide fine-grained vulnerability analysis capability. In addition, many of the fault injection tools still need to change the neural network models and make the fault injection closely coupled with normal neural network processing, which further complicates the use of the fault injection tools and slows down the fault simulation. In this work, we propose MRFI, a highly configurable multi-resolution fault injection tool for deep neural networks. It enables users to modify an independent fault configuration file rather than neural network models for the fault injection and vulnerability analysis. Particularly, it integrates extensive fault analysis functionalities from different perspectives and enables multi-resolution investigation of the vulnerability of neural networks. In addition, it does not modify the major neural network computing framework of PyTorch. Hence, it allows parallel processing on GPUs naturally and exhibits fast fault simulation according to our experiments.

Abstract: Cooperative multi-agent reinforcement learning (MARL) for navigation enables agents to cooperate to achieve their navigation goals. Using emergent communication, agents learn a communication protocol to coordinate and share information that is needed to achieve their navigation tasks. In emergent communication, symbols with no pre-specified usage rules are exchanged, in which the meaning and syntax emerge through training. Learning a navigation policy along with a communication protocol in a MARL environment is highly complex due to the huge state space to be explored. To cope with this complexity, this work proposes a novel neural network architecture, for jointly learning an adaptive state space abstraction and a communication protocol among agents participating in navigation tasks. The goal is to come up with an adaptive abstractor that significantly reduces the size of the state space to be explored, without degradation in the policy performance. Simulation results show that the proposed method reaches a better policy, in terms of achievable rewards, resulting in fewer training iterations compared to the case where raw states or fixed state abstraction are used. Moreover, it is shown that a communication protocol emerges during training which enables the agents to learn better policies within fewer training iterations.

Abstract: The concentration of power in a few digital technology companies has become a subject of increasing interest in both academic and non-academic discussions. One of the most noteworthy contributions to the debate is Lina Khan's Amazon's Antitrust Paradox. In this work, Khan contends that Amazon has systematically exerted its dominance in online retail to eliminate competitors and subsequently charge above-market prices. This work contributed to Khan's appointment as the chair of the US Federal Trade Commission (FTC), one of the most influential antitrust organizations. Today, several ongoing antitrust lawsuits in the US and Europe involve major technology companies like Apple, Google/Alphabet, and Facebook/Meta. In the realm of generative AI, we are once again witnessing the same companies taking the lead in technological advancements, leaving little room for others to compete. This article examines the market dominance of these corporations in the technology stack behind generative AI from an antitrust law perspective.

Abstract: In conventional supervised classification, true labels are required for individual instances. However, it could be prohibitive to collect the true labels for individual instances, due to privacy concerns or unaffordable annotation costs. This motivates the study on classification from aggregate observations (CFAO), where the supervision is provided to groups of instances, instead of individual instances. CFAO is a generalized learning framework that contains various learning problems, such as multiple-instance learning and learning from label proportions. The goal of this paper is to present a novel universal method of CFAO, which holds an unbiased estimator of the classification risk for arbitrary losses -- previous research failed to achieve this goal. Practically, our method works by weighing the importance of each label for each instance in the group, which provides purified supervision for the classifier to learn. Theoretically, our proposed method not only guarantees the risk consistency due to the unbiased risk estimator but also can be compatible with arbitrary losses. Extensive experiments on various problems of CFAO demonstrate the superiority of our proposed method.

Abstract: Contrastive learning methods have attracted considerable attention due to their remarkable success in analyzing graph-structured data. Inspired by the success of contrastive learning, we propose a novel framework for contrastive disentangled learning on graphs, employing a disentangled graph encoder and two carefully crafted self-supervision signals. Specifically, we introduce a disentangled graph encoder to enforce the framework to distinguish various latent factors corresponding to underlying semantic information and learn the disentangled node embeddings. Moreover, to overcome the heavy reliance on labels, we design two self-supervision signals, namely node specificity and channel independence, which capture informative knowledge without the need for labeled data, thereby guiding the automatic disentanglement of nodes. Finally, we perform node classification tasks on three citation networks by using the disentangled node embeddings, and the relevant analysis is provided. Experimental results validate the effectiveness of the proposed framework compared with various baselines.

Abstract: Large graphs are present in a variety of domains, including social networks, civil infrastructure, and the physical sciences to name a few. Graph generation is similarly widespread, with applications in drug discovery, network analysis and synthetic datasets among others. While GNN (Graph Neural Network) models have been applied in these domains their high in-memory costs restrict them to small graphs. Conversely less costly rule-based methods struggle to reproduce complex structures. We propose HIGGS (Hierarchical Generation of Graphs) as a model-agnostic framework of producing large graphs with realistic local structures. HIGGS uses GNN models with conditional generation capabilities to sample graphs in hierarchies of resolution. As a result HIGGS has the capacity to extend the scale of generated graphs from a given GNN model by quadratic order. As a demonstration we implement HIGGS using DiGress, a recent graph-diffusion model, including a novel edge-predictive-diffusion variant edge-DiGress. We use this implementation to generate categorically attributed graphs with tens of thousands of nodes. These HIGGS generated graphs are far larger than any previously produced using GNNs. Despite this jump in scale we demonstrate that the graphs produced by HIGGS are, on the local scale, more realistic than those from the rule-based model BTER.

Abstract: Deep learning (DL) is characterised by its dynamic nature, with new deep neural network (DNN) architectures and approaches emerging every few years, driving the field's advancement. At the same time, the ever-increasing use of mobile devices (MDs) has resulted in a surge of DNN-based mobile applications. Although traditional architectures, like CNNs and RNNs, have been successfully integrated into MDs, this is not the case for Transformers, a relatively new model family that has achieved new levels of accuracy across AI tasks, but poses significant computational challenges. In this work, we aim to make steps towards bridging this gap by examining the current state of Transformers' on-device execution. To this end, we construct a benchmark of representative models and thoroughly evaluate their performance across MDs with different computational capabilities. Our experimental results show that Transformers are not accelerator-friendly and indicate the need for software and hardware optimisations to achieve efficient deployment.

Abstract: Neural networks (NNs) that exploit strong inductive biases based on physical laws and symmetries have shown remarkable success in learning the dynamics of physical systems directly from their trajectory. However, these works focus only on the systems that follow deterministic dynamics, for instance, Newtonian or Hamiltonian dynamics. Here, we propose a framework, namely Brownian graph neural networks (BROGNET), combining stochastic differential equations (SDEs) and GNNs to learn Brownian dynamics directly from the trajectory. We theoretically show that BROGNET conserves the linear momentum of the system, which in turn, provides superior performance on learning dynamics as revealed empirically. We demonstrate this approach on several systems, namely, linear spring, linear spring with binary particle types, and non-linear spring systems, all following Brownian dynamics at finite temperatures. We show that BROGNET significantly outperforms proposed baselines across all the benchmarked Brownian systems. In addition, we demonstrate zero-shot generalizability of BROGNET to simulate unseen system sizes that are two orders of magnitude larger and to different temperatures than those used during training. Altogether, our study contributes to advancing the understanding of the intricate dynamics of Brownian motion and demonstrates the effectiveness of graph neural networks in modeling such complex systems.

Abstract: In a broad class of reinforcement learning applications, stochastic rewards have heavy-tailed distributions, which lead to infinite second-order moments for stochastic (semi)gradients in policy evaluation and direct policy optimization. In such instances, the existing RL methods may fail miserably due to frequent statistical outliers. In this work, we establish that temporal difference (TD) learning with a dynamic gradient clipping mechanism, and correspondingly operated natural actor-critic (NAC), can be provably robustified against heavy-tailed reward distributions. It is shown in the framework of linear function approximation that a favorable tradeoff between bias and variability of the stochastic gradients can be achieved with this dynamic gradient clipping mechanism. In particular, we prove that robust versions of TD learning achieve sample complexities of order $\mathcal{O}(\varepsilon^{-\frac{1}{p}})$ and $\mathcal{O}(\varepsilon^{-1-\frac{1}{p}})$ with and without the full-rank assumption on the feature matrix, respectively, under heavy-tailed rewards with finite moments of order $(1+p)$ for some $p\in(0,1]$, both in expectation and with high probability. We show that a robust variant of NAC based on Robust TD learning achieves $\tilde{\mathcal{O}}(\varepsilon^{-4-\frac{2}{p}})$ sample complexity. We corroborate our theoretical results with numerical experiments.

Abstract: When machine learning is outsourced to a rational agent, conflicts of interest might arise and severely impact predictive performance. In this work, we propose a theoretical framework for incentive-aware delegation of machine learning tasks. We model delegation as a principal-agent game, in which accurate learning can be incentivized by the principal using performance-based contracts. Adapting the economic theory of contract design to this setting, we define budget-optimal contracts and prove they take a simple threshold form under reasonable assumptions. In the binary-action case, the optimality of such contracts is shown to be equivalent to the classic Neyman-Pearson lemma, establishing a formal connection between contract design and statistical hypothesis testing. Empirically, we demonstrate that budget-optimal contracts can be constructed using small-scale data, leveraging recent advances in the study of learning curves and scaling laws. Performance and economic outcomes are evaluated using synthetic and real-world classification tasks.

Abstract: This paper proposes a new framework for learning a rule ensemble model that is both accurate and interpretable. A rule ensemble is an interpretable model based on the linear combination of weighted rules. In practice, we often face the trade-off between the accuracy and interpretability of rule ensembles. That is, a rule ensemble needs to include a sufficiently large number of weighted rules to maintain its accuracy, which harms its interpretability for human users. To avoid this trade-off and learn an interpretable rule ensemble without degrading accuracy, we introduce a new concept of interpretability, named local interpretability, which is evaluated by the total number of rules necessary to express individual predictions made by the model, rather than to express the model itself. Then, we propose a regularizer that promotes local interpretability and develop an efficient algorithm for learning a rule ensemble with the proposed regularizer by coordinate descent with local search. Experimental results demonstrated that our method learns rule ensembles that can explain individual predictions with fewer rules than the existing methods, including RuleFit, while maintaining comparable accuracy.

Abstract: While deep reinforcement learning (RL) agents outperform humans on an increasing number of tasks, training them requires data equivalent to decades of human gameplay. Recent hierarchical RL methods have increased sample efficiency by incorporating information inherent to the structure of the decision problem but at the cost of having to discover or use human-annotated sub-goals that guide the learning process. We show that intentions of human players, i.e. the precursor of goal-oriented decisions, can be robustly predicted from eye gaze even for the long-horizon sparse rewards task of Montezuma's Revenge - one of the most challenging RL tasks in the Atari2600 game suite. We propose Int-HRL: Hierarchical RL with intention-based sub-goals that are inferred from human eye gaze. Our novel sub-goal extraction pipeline is fully automatic and replaces the need for manual sub-goal annotation by human experts. Our evaluations show that replacing hand-crafted sub-goals with automatically extracted intentions leads to a HRL agent that is significantly more sample efficient than previous methods.

Abstract: In this work, we generalize the problem of learning through interaction in a POMDP by accounting for eventual additional information available at training time. First, we introduce the informed POMDP, a new learning paradigm offering a clear distinction between the training information and the execution observation. Next, we propose an objective for learning a sufficient statistic from the history for the optimal control that leverages this information. We then show that this informed objective consists of learning an environment model from which we can sample latent trajectories. Finally, we show for the Dreamer algorithm that the convergence speed of the policies is sometimes greatly improved on several environments by using this informed environment model. Those results and the simplicity of the proposed adaptation advocate for a systematic consideration of eventual additional information when learning in a POMDP using model-based RL.

Abstract: Self-supervised learning, or SSL, holds the key to expanding the usage of machine learning in real-world tasks by alleviating heavy human supervision. Contrastive learning and its varieties have been SSL strategies in various fields. We use margins as a stepping stone for understanding how contrastive learning works at a deeper level and providing potential directions to improve representation learning. Through gradient analysis, we found that margins scale gradients in three different ways: emphasizing positive samples, de-emphasizing positive samples when angles of positive samples are wide, and attenuating the diminishing gradients as the estimated probability approaches the target probability. We separately analyze each and provide possible directions for improving SSL frameworks. Our experimental results demonstrate that these properties can contribute to acquiring better representations, which can enhance performance in both seen and unseen datasets.

Abstract: Generative, pre-trained transformers (GPTs, a.k.a. "Foundation Models") have reshaped natural language processing (NLP) through their versatility in diverse downstream tasks. However, their potential extends far beyond NLP. This paper provides a software utility to help realize this potential, extending the applicability of GPTs to continuous-time sequences of complex events with internal dependencies, such as medical record datasets. Despite their potential, the adoption of foundation models in these domains has been hampered by the lack of suitable tools for model construction and evaluation. To bridge this gap, we introduce Event Stream GPT (ESGPT), an open-source library designed to streamline the end-to-end process for building GPTs for continuous-time event sequences. ESGPT allows users to (1) build flexible, foundation-model scale input datasets by specifying only a minimal configuration file, (2) leverage a Hugging Face compatible modeling API for GPTs over this modality that incorporates intra-event causal dependency structures and autoregressive generation capabilities, and (3) evaluate models via standardized processes that can assess few and even zero-shot performance of pre-trained models on user-specified fine-tuning tasks.

Abstract: We introduce IMP-MARL, an open-source suite of multi-agent reinforcement learning (MARL) environments for large-scale Infrastructure Management Planning (IMP), offering a platform for benchmarking the scalability of cooperative MARL methods in real-world engineering applications. In IMP, a multi-component engineering system is subject to a risk of failure due to its components' damage condition. Specifically, each agent plans inspections and repairs for a specific system component, aiming to minimise maintenance costs while cooperating to minimise system failure risk. With IMP-MARL, we release several environments including one related to offshore wind structural systems, in an effort to meet today's needs to improve management strategies to support sustainable and reliable energy systems. Supported by IMP practical engineering environments featuring up to 100 agents, we conduct a benchmark campaign, where the scalability and performance of state-of-the-art cooperative MARL methods are compared against expert-based heuristic policies. The results reveal that centralised training with decentralised execution methods scale better with the number of agents than fully centralised or decentralised RL approaches, while also outperforming expert-based heuristic policies in most IMP environments. Based on our findings, we additionally outline remaining cooperation and scalability challenges that future MARL methods should still address. Through IMP-MARL, we encourage the implementation of new environments and the further development of MARL methods.

Abstract: Despite deep learning has achieved great success, it often relies on a large amount of training data with accurate labels, which are expensive and time-consuming to collect. A prominent direction to reduce the cost is to learn with noisy labels, which are ubiquitous in the real-world applications. A critical challenge for such a learning task is to reduce the effect of network memorization on the falsely-labeled data. In this work, we propose an iterative selection approach based on the Weibull mixture model, which identifies clean data by considering the overall learning dynamics of each data instance. In contrast to the previous small-loss heuristics, we leverage the observation that deep network is easy to memorize and hard to forget clean data. In particular, we measure the difficulty of memorization and forgetting for each instance via the transition times between being misclassified and being memorized in training, and integrate them into a novel metric for selection. Based on the proposed metric, we retain a subset of identified clean data and repeat the selection procedure to iteratively refine the clean subset, which is finally used for model training. To validate our method, we perform extensive experiments on synthetic noisy datasets and real-world web data, and our strategy outperforms existing noisy-label learning methods.

Abstract: This paper proposes a set of simple adaptations to transform standard message-passing Graph Neural Networks (GNN) into provably powerful directed multigraph neural networks. The adaptations include multigraph port numbering, ego IDs, and reverse message passing. We prove that the combination of these theoretically enables the detection of any directed subgraph pattern. To validate the effectiveness of our proposed adaptations in practice, we conduct experiments on synthetic subgraph detection tasks, which demonstrate outstanding performance with almost perfect results. Moreover, we apply our proposed adaptations to two financial crime analysis tasks. We observe dramatic improvements in detecting money laundering transactions, improving the minority-class F1 score of a standard message-passing GNN by up to 45%, and clearly outperforming tree-based and GNN baselines. Similarly impressive results are observed on a real-world phishing detection dataset, boosting a standard GNN's F1 score by over 15% and outperforming all baselines.

Abstract: Gaussian processes are a powerful framework for quantifying uncertainty and for sequential decision-making but are limited by the requirement of solving linear systems. In general, this has a cubic cost in dataset size and is sensitive to conditioning. We explore stochastic gradient algorithms as a computationally efficient method of approximately solving these linear systems: we develop low-variance optimization objectives for sampling from the posterior and extend these to inducing points. Counterintuitively, stochastic gradient descent often produces accurate predictions, even in cases where it does not converge quickly to the optimum. We explain this through a spectral characterization of the implicit bias from non-convergence. We show that stochastic gradient descent produces predictive distributions close to the true posterior both in regions with sufficient data coverage, and in regions sufficiently far away from the data. Experimentally, stochastic gradient descent achieves state-of-the-art performance on sufficiently large-scale or ill-conditioned regression tasks. Its uncertainty estimates match the performance of significantly more expensive baselines on a large-scale Bayesian~optimization~task.

Abstract: Machine learning applications cover a wide range of predictive tasks in which tabular datasets play a significant role. However, although they often address similar problems, tabular datasets are typically treated as standalone tasks. The possibilities of using previously solved problems are limited due to the lack of structured contextual information about their features and the lack of understanding of the relations between them. To overcome this limitation, we propose a new approach called Semantic Feature Net (SeFNet), capturing the semantic meaning of the analyzed tabular features. By leveraging existing ontologies and domain knowledge, SeFNet opens up new opportunities for sharing insights between diverse predictive tasks. One such opportunity is the Dataset Ontology-based Semantic Similarity (DOSS) measure, which quantifies the similarity between datasets using relations across their features. In this paper, we present an example of SeFNet prepared for a collection of predictive tasks in healthcare, with the features' relations derived from the SNOMED-CT ontology. The proposed SeFNet framework and the accompanying DOSS measure address the issue of limited contextual information in tabular datasets. By incorporating domain knowledge and establishing semantic relations between features, we enhance the potential for meta-learning and enable valuable insights to be shared across different predictive tasks.

Abstract: Federated learning has gained popularity for distributed learning without aggregating sensitive data from clients. But meanwhile, the distributed and isolated nature of data isolation may be complicated by data quality, making it more vulnerable to noisy labels. Many efforts exist to defend against the negative impacts of noisy labels in centralized or federated settings. However, there is a lack of a benchmark that comprehensively considers the impact of noisy labels in a wide variety of typical FL settings. In this work, we serve the first standardized benchmark that can help researchers fully explore potential federated noisy settings. Also, we conduct comprehensive experiments to explore the characteristics of these data settings and unravel challenging scenarios on the federated noisy label learning, which may guide method development in the future. We highlight the 20 basic settings for more than 5 datasets proposed in our benchmark and standardized simulation pipeline for federated noisy label learning. We hope this benchmark can facilitate idea verification in federated learning with noisy labels. \texttt{FedNoisy} is available at \codeword{https://github.com/SMILELab-FL/FedNoisy}.

Abstract: In this study, we focus on the development and implementation of a comprehensive ensemble of numerical time series forecasting models, collectively referred to as the Group of Numerical Time Series Prediction Model (G-NM). This inclusive set comprises traditional models such as Autoregressive Integrated Moving Average (ARIMA), Holt-Winters' method, and Support Vector Regression (SVR), in addition to modern neural network models including Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM). G-NM is explicitly constructed to augment our predictive capabilities related to patterns and trends inherent in complex natural phenomena. By utilizing time series data relevant to these events, G-NM facilitates the prediction of such phenomena over extended periods. The primary objective of this research is to both advance our understanding of such occurrences and to significantly enhance the accuracy of our forecasts. G-NM encapsulates both linear and non-linear dependencies, seasonalities, and trends present in time series data. Each of these models contributes distinct strengths, from ARIMA's resilience in handling linear trends and seasonality, SVR's proficiency in capturing non-linear patterns, to LSTM's adaptability in modeling various components of time series data. Through the exploitation of the G-NM potential, we strive to advance the state-of-the-art in large-scale time series forecasting models. We anticipate that this research will represent a significant stepping stone in our ongoing endeavor to comprehend and forecast the complex events that constitute the natural world.

Abstract: It is often advantageous to train models on a subset of the available train examples, because the examples are of variable quality or because one would like to train with fewer examples, without sacrificing performance. We present Gradient Information Optimization (GIO), a scalable, task-agnostic approach to this data selection problem that requires only a small set of (unlabeled) examples representing a target distribution. GIO begins from a natural, information-theoretic objective that is intractable in practice. Our contribution is in showing that it can be made highly scalable through a simple relaxation of the objective and a highly efficient implementation. In experiments with machine translation, spelling correction, and image recognition, we show that GIO delivers outstanding results with very small train sets. These findings are robust to different representation models and hyperparameters for GIO itself. GIO is task- and domain-agnostic and can be applied out-of-the-box to new datasets and domains.

Abstract: We study the implicit bias of batch normalization trained by gradient descent. We show that when learning a linear model with batch normalization for binary classification, gradient descent converges to a uniform margin classifier on the training data with an $\exp(-\Omega(\log^2 t))$ convergence rate. This distinguishes linear models with batch normalization from those without batch normalization in terms of both the type of implicit bias and the convergence rate. We further extend our result to a class of two-layer, single-filter linear convolutional neural networks, and show that batch normalization has an implicit bias towards a patch-wise uniform margin. Based on two examples, we demonstrate that patch-wise uniform margin classifiers can outperform the maximum margin classifiers in certain learning problems. Our results contribute to a better theoretical understanding of batch normalization.

Abstract: Structure-based models in the molecular sciences can be highly sensitive to input geometries and give predictions with large variance under subtle coordinate perturbations. We present an approach to mitigate this failure mode by generating conformations that explicitly minimize uncertainty in a predictive model. To achieve this, we compute differentiable estimates of aleatoric \textit{and} epistemic uncertainties directly from learned embeddings. We then train an optimizer that iteratively samples embeddings to reduce these uncertainties according to their gradients. As our predictive model is constructed as a variational autoencoder, the new embeddings can be decoded to their corresponding inputs, which we call \textit{MoleCLUEs}, or (molecular) counterfactual latent uncertainty explanations \citep{antoran2020getting}. We provide results of our algorithm for the task of predicting drug properties with maximum confidence as well as analysis of the differentiable structure simulations.

Abstract: We propose RoTaR, a row-based table representation learning method, to address the efficiency and scalability issues faced by existing table representation learning methods. The key idea of RoTaR is to generate query-agnostic row representations that could be re-used via query-specific aggregation. In addition to the row-based architecture, we introduce several techniques: cell-aware position embedding, teacher-student training paradigm, and selective backward to improve the performance of RoTaR model.

Abstract: In the last decades, the capacity to generate large amounts of data in science and engineering applications has been growing steadily. Meanwhile, the progress in machine learning has turned it into a suitable tool to process and utilise the available data. Nonetheless, many relevant scientific and engineering problems present challenges where current machine learning methods cannot yet efficiently leverage the available data and resources. For example, in scientific discovery, we are often faced with the problem of exploring very large, high-dimensional spaces, where querying a high fidelity, black-box objective function is very expensive. Progress in machine learning methods that can efficiently tackle such problems would help accelerate currently crucial areas such as drug and materials discovery. In this paper, we propose the use of GFlowNets for multi-fidelity active learning, where multiple approximations of the black-box function are available at lower fidelity and cost. GFlowNets are recently proposed methods for amortised probabilistic inference that have proven efficient for exploring large, high-dimensional spaces and can hence be practical in the multi-fidelity setting too. Here, we describe our algorithm for multi-fidelity active learning with GFlowNets and evaluate its performance in both well-studied synthetic tasks and practically relevant applications of molecular discovery. Our results show that multi-fidelity active learning with GFlowNets can efficiently leverage the availability of multiple oracles with different costs and fidelities to accelerate scientific discovery and engineering design.

Abstract: With the increase in the scale of Deep Learning (DL) training workloads in terms of compute resources and time consumption, the likelihood of encountering in-training failures rises substantially, leading to lost work and resource wastage. Such failures are typically offset by a checkpointing mechanism, which comes at the cost of storage and network bandwidth overhead. State-of-the-art approaches involve lossy model compression mechanisms, which induce a tradeoff between the resulting model quality (accuracy) and compression ratio. Delta compression is then also used to further reduce the overhead by only storing the difference between consecutive checkpoints. We make a key enabling observation that the sensitivity of model weights to compression varies during training, and different weights benefit from different quantization levels (ranging from retaining full precision to pruning). We propose (1) a non-uniform quantization scheme that leverages this variation, (2) an efficient search mechanism to dynamically adjust to the best quantization configurations, and (3) a quantization-aware delta compression mechanism that rearranges weights to minimize checkpoint differences, thereby maximizing compression. We instantiate these contributions in DynaQuant - a framework for DL workload checkpoint compression. Our experiments show that DynaQuant consistently achieves better tradeoff between accuracy and compression ratios compared to prior works, enabling a compression ratio up to 39x and withstanding up to 10 restores with negligible accuracy impact for fault-tolerant training. DynaQuant achieves at least an order of magnitude reduction in checkpoint storage overhead for training failure recovery as well as transfer learning use cases without any loss of accuracy

Abstract: Reinforcement learning (RL) has emerged as a powerful paradigm for fine-tuning Large Language Models (LLMs) for conditional text generation. In particular, recent LLMs such as ChatGPT and GPT-4 can engage in fluent conversations with users by incorporating RL and feedback from humans. Inspired by learning-to-search algorithms and capitalizing on key properties of text generation, we seek to investigate reinforcement learning algorithms beyond general purpose algorithms such as Proximal policy optimization (PPO). In particular, we extend RL algorithms to allow them to interact with a dynamic black-box guide LLM such as GPT-3 and propose RL with guided feedback (RLGF), a suite of RL algorithms for LLM fine-tuning. We experiment on the IMDB positive review and CommonGen text generation task from the GRUE benchmark. We show that our RL algorithms achieve higher performance than supervised learning (SL) and default PPO baselines, demonstrating the benefit of interaction with the guide LLM. On CommonGen, we not only outperform our SL baselines but also improve beyond PPO across a variety of lexical and semantic metrics beyond the one we optimized for. Notably, on the IMDB dataset, we show that our GPT-2 based policy outperforms the zero-shot GPT-3 oracle, indicating that our algorithms can learn from a powerful, black-box GPT-3 oracle with a simpler, cheaper, and publicly available GPT-2 model while gaining performance.

Abstract: We constructively prove that every deep ReLU network can be rewritten as a functionally identical three-layer network with weights valued in the extended reals. Based on this proof, we provide an algorithm that, given a deep ReLU network, finds the explicit weights of the corresponding shallow network. The resulting shallow network is transparent and used to generate explanations of the model s behaviour.

Abstract: Many brain-computer interfaces make use of brain signals that are elicited in response to a visual, auditory or tactile stimulus, so-called event-related potentials (ERPs). In visual ERP speller applications, sets of letters shown on a screen are flashed randomly, and the participant attends to the target letter they want to spell. When this letter flashes, the resulting ERP is different compared to when any other non-target letter flashes. We propose a new unsupervised approach to detect this attended letter. In each trial, for every available letter our approach makes the hypothesis that it is in fact the attended letter, and calculates the ERPs based on each of these hypotheses. We leverage the fact that only the true hypothesis produces the largest difference between the class means. Note that this unsupervised method does not require any changes to the underlying experimental paradigm and therefore can be employed in almost any ERP-based setup. To deal with limited data, we use a block-Toeplitz regularized covariance matrix that models the background activity. We implemented the proposed novel unsupervised mean-difference maximization (UMM) method and evaluated it in offline replays of brain-computer interface visual speller datasets. For a dataset that used 16 flashes per symbol per trial, UMM correctly classifies 3651 out of 3654 letters ($99.92\,\%$) across 25 participants. In another dataset with fewer and shorter trials, 7344 out of 7383 letters ($99.47\,\%$) are classified correctly across 54 participants with two sessions each. Even in more challenging datasets obtained from patients with amyotrophic lateral sclerosis ($77.86\,\%$) or when using auditory ERPs ($82.52\,\%$), the obtained classification rates obtained by UMM are competitive. In addition, UMM provides stable confidence measures which can be used to monitor convergence.

Abstract: For safety and robustness of AI systems, we introduce topological parallax as a theoretical and computational tool that compares a trained model to a reference dataset to determine whether they have similar multiscale geometric structure. Our proofs and examples show that this geometric similarity between dataset and model is essential to trustworthy interpolation and perturbation, and we conjecture that this new concept will add value to the current debate regarding the unclear relationship between overfitting and generalization in applications of deep-learning. In typical DNN applications, an explicit geometric description of the model is impossible, but parallax can estimate topological features (components, cycles, voids, etc.) in the model by examining the effect on the Rips complex of geodesic distortions using the reference dataset. Thus, parallax indicates whether the model shares similar multiscale geometric features with the dataset. Parallax presents theoretically via topological data analysis [TDA] as a bi-filtered persistence module, and the key properties of this module are stable under perturbation of the reference dataset.

Abstract: This study investigates the interplay among social demographics, built environment characteristics, and environmental hazard exposure features in determining community level cancer prevalence. Utilizing data from five Metropolitan Statistical Areas in the United States: Chicago, Dallas, Houston, Los Angeles, and New York, the study implemented an XGBoost machine learning model to predict the extent of cancer prevalence and evaluate the importance of different features. Our model demonstrates reliable performance, with results indicating that age, minority status, and population density are among the most influential factors in cancer prevalence. We further explore urban development and design strategies that could mitigate cancer prevalence, focusing on green space, developed areas, and total emissions. Through a series of experimental evaluations based on causal inference, the results show that increasing green space and reducing developed areas and total emissions could alleviate cancer prevalence. The study and findings contribute to a better understanding of the interplay among urban features and community health and also show the value of interpretable machine learning models for integrated urban design to promote public health. The findings also provide actionable insights for urban planning and design, emphasizing the need for a multifaceted approach to addressing urban health disparities through integrated urban design strategies.

Abstract: Understanding the effect of a feature vector $x \in \mathbb{R}^d$ on the response value (label) $y \in \mathbb{R}$ is the cornerstone of many statistical learning problems. Ideally, it is desired to understand how a set of collected features combine together and influence the response value, but this problem is notoriously difficult, due to the high-dimensionality of data and limited number of labeled data points, among many others. In this work, we take a new perspective on this problem, and we study the question of assessing the difference of influence that the two given features have on the response value. We first propose a notion of closeness for the influence of features, and show that our definition recovers the familiar notion of the magnitude of coefficients in the parametric model. We then propose a novel method to test for the closeness of influence in general model-free supervised learning problems. Our proposed test can be used with finite number of samples with control on type I error rate, no matter the ground truth conditional law $\mathcal{L}(Y |X)$. We analyze the power of our test for two general learning problems i) linear regression, and ii) binary classification under mixture of Gaussian models, and show that under the proper choice of score function, an internal component of our test, with sufficient number of samples will achieve full statistical power. We evaluate our findings through extensive numerical simulations, specifically we adopt the datamodel framework (Ilyas, et al., 2022) for CIFAR-10 dataset to identify pairs of training samples with different influence on the trained model via optional black box training mechanisms.

Abstract: Transmit power control (TPC) is a key mechanism for managing interference, energy utilization, and connectivity in wireless systems. In this paper, we propose a simple low-complexity TPC algorithm based on the deep unfolding of the iterative projected gradient descent (PGD) algorithm into layers of a deep neural network and learning the step-size parameters. An unsupervised learning method with either online learning or offline pretraining is applied for optimizing the weights of the DNN. Performance evaluation in dense device-to-device (D2D) communication scenarios showed that the proposed method can achieve better performance than the iterative algorithm with more than a factor of 2 lower number of iterations.

Abstract: Data heterogeneity is one of the most challenging issues in federated learning, which motivates a variety of approaches to learn personalized models for participating clients. One such approach in deep neural networks based tasks is employing a shared feature representation and learning a customized classifier head for each client. However, previous works do not utilize the global knowledge during local representation learning and also neglect the fine-grained collaboration between local classifier heads, which limit the model generalization ability. In this work, we conduct explicit local-global feature alignment by leveraging global semantic knowledge for learning a better representation. Moreover, we quantify the benefit of classifier combination for each client as a function of the combining weights and derive an optimization problem for estimating optimal weights. Finally, extensive evaluation results on benchmark datasets with various heterogeneous data scenarios demonstrate the effectiveness of our proposed method. Code is available at https://github.com/JianXu95/FedPAC

Abstract: Reinforcement Learning (RL) models have continually evolved to navigate the exploration - exploitation trade-off in uncertain Markov Decision Processes (MDPs). In this study, I leverage the principles of stochastic thermodynamics and system dynamics to explore reward shaping via diffusion processes. This provides an elegant framework as a way to think about exploration-exploitation trade-off. This article sheds light on relationships between information entropy, stochastic system dynamics, and their influences on entropy production. This exploration allows us to construct a dual-pronged framework that can be interpreted as either a maximum entropy program for deriving efficient policies or a modified cost optimization program accounting for informational costs and benefits. This work presents a novel perspective on the physical nature of information and its implications for online learning in MDPs, consequently providing a better understanding of information-oriented formulations in RL.

Abstract: It has become standard to explain neural network latent spaces with attraction/repulsion dimensionality reduction (ARDR) methods like tSNE and UMAP. This relies on the premise that structure in the 2D representation is consistent with the structure in the model's latent space. However, this is an unproven assumption -- we are unaware of any convergence guarantees for ARDR algorithms. We work on closing this question by relating ARDR methods to classical dimensionality reduction techniques. Specifically, we show that one can fully recover a PCA embedding by applying attractions and repulsions onto a randomly initialized dataset. We also show that, with a small change, Locally Linear Embeddings (LLE) can reproduce ARDR embeddings. Finally, we formalize a series of conjectures that, if true, would allow one to attribute structure in the 2D embedding back to the input distribution.

Abstract: In recent years, deep learning has made remarkable progress in a wide range of domains, with a particularly notable impact on natural language processing tasks. One of the challenges associated with training deep neural networks is the need for large amounts of computational resources and time. In this paper, we present Deep Fusion, an efficient approach to network training that leverages pre-trained initializations of smaller networks. % We show that Deep Fusion accelerates the training process, reduces computational requirements, and leads to improved generalization performance on a variety of NLP tasks and T5 model sizes. % Our experiments demonstrate that Deep Fusion is a practical and effective approach to reduce the training time and resource consumption while maintaining, or even surpassing, the performance of traditional training methods.

Abstract: A survival dataset describes a set of instances (e.g. patients) and provides, for each, either the time until an event (e.g. death), or the censoring time (e.g. when lost to follow-up - which is a lower bound on the time until the event). We consider the challenge of survival prediction: learning, from such data, a predictive model that can produce an individual survival distribution for a novel instance. Many contemporary methods of survival prediction implicitly assume that the event and censoring distributions are independent conditional on the instance's covariates - a strong assumption that is difficult to verify (as we observe only one outcome for each instance) and which can induce significant bias when it does not hold. This paper presents a parametric model of survival that extends modern non-linear survival analysis by relaxing the assumption of conditional independence. On synthetic and semi-synthetic data, our approach significantly improves estimates of survival distributions compared to the standard that assumes conditional independence in the data.

Abstract: Federated learning (FL) is a common and practical framework for learning a machine model in a decentralized fashion. A primary motivation behind this decentralized approach is data privacy, ensuring that the learner never sees the data of each local source itself. Federated learning then comes with two majors challenges: one is handling potentially complex model updates between a server and a large number of data sources; the other is that de-centralization may, in fact, be insufficient for privacy, as the local updates themselves can reveal information about the sources' data. To address these issues, we consider an approach to federated learning that combines quantization and differential privacy. Absent privacy, Federated Learning often relies on quantization to reduce communication complexity. We build upon this approach and develop a new algorithm called the \textbf{R}andomized \textbf{Q}uantization \textbf{M}echanism (RQM), which obtains privacy through a two-levels of randomization. More precisely, we randomly sub-sample feasible quantization levels, then employ a randomized rounding procedure using these sub-sampled discrete levels. We are able to establish that our results preserve ``Renyi differential privacy'' (Renyi DP). We empirically study the performance of our algorithm and demonstrate that compared to previous work it yields improved privacy-accuracy trade-offs for DP federated learning. To the best of our knowledge, this is the first study that solely relies on randomized quantization without incorporating explicit discrete noise to achieve Renyi DP guarantees in Federated Learning systems.

Abstract: Certifying the robustness of a graph-based machine learning model poses a critical challenge for safety. Current robustness certificates for graph classifiers guarantee output invariance with respect to the total number of node pair flips (edge addition or edge deletion), which amounts to an $l_{0}$ ball centred on the adjacency matrix. Although theoretically attractive, this type of isotropic structural noise can be too restrictive in practical scenarios where some node pairs are more critical than others in determining the classifier's output. The certificate, in this case, gives a pessimistic depiction of the robustness of the graph model. To tackle this issue, we develop a randomised smoothing method based on adding an anisotropic noise distribution to the input graph structure. We show that our process generates structural-aware certificates for our classifiers, whereby the magnitude of robustness certificates can vary across different pre-defined structures of the graph. We demonstrate the benefits of these certificates in both synthetic and real-world experiments.

Abstract: The ensemble method is a promising way to mitigate the overestimation issue in Q-learning, where multiple function approximators are used to estimate the action values. It is known that the estimation bias hinges heavily on the ensemble size (i.e., the number of Q-function approximators used in the target), and that determining the `right' ensemble size is highly nontrivial, because of the time-varying nature of the function approximation errors during the learning process. To tackle this challenge, we first derive an upper bound and a lower bound on the estimation bias, based on which the ensemble size is adapted to drive the bias to be nearly zero, thereby coping with the impact of the time-varying approximation errors accordingly. Motivated by the theoretic findings, we advocate that the ensemble method can be combined with Model Identification Adaptive Control (MIAC) for effective ensemble size adaptation. Specifically, we devise Adaptive Ensemble Q-learning (AdaEQ), a generalized ensemble method with two key steps: (a) approximation error characterization which serves as the feedback for flexibly controlling the ensemble size, and (b) ensemble size adaptation tailored towards minimizing the estimation bias. Extensive experiments are carried out to show that AdaEQ can improve the learning performance than the existing methods for the MuJoCo benchmark.

Abstract: Understanding the optimization dynamics of neural networks is necessary for closing the gap between theory and practice. Stochastic first-order optimization algorithms are known to efficiently locate favorable minima in deep neural networks. This efficiency, however, contrasts with the non-convex and seemingly complex structure of neural loss landscapes. In this study, we delve into the fundamental geometric properties of sampled gradients along optimization paths. We focus on two key quantities, which appear in the restricted secant inequality and error bound. Both hold high significance for first-order optimization. Our analysis reveals that these quantities exhibit predictable, consistent behavior throughout training, despite the stochasticity induced by sampling minibatches. Our findings suggest that not only do optimization trajectories never encounter significant obstacles, but they also maintain stable dynamics during the majority of training. These observed properties are sufficiently expressive to theoretically guarantee linear convergence and prescribe learning rate schedules mirroring empirical practices. We conduct our experiments on image classification, semantic segmentation and language modeling across different batch sizes, network architectures, datasets, optimizers, and initialization seeds. We discuss the impact of each factor. Our work provides novel insights into the properties of neural network loss functions, and opens the door to theoretical frameworks more relevant to prevalent practice.

Abstract: The accurate modeling of dynamics in interactive environments is critical for successful long-range prediction. Such a capability could advance Reinforcement Learning (RL) and Planning algorithms, but achieving it is challenging. Inaccuracies in model estimates can compound, resulting in increased errors over long horizons. We approach this problem from the lens of Koopman theory, where the nonlinear dynamics of the environment can be linearized in a high-dimensional latent space. This allows us to efficiently parallelize the sequential problem of long-range prediction using convolution, while accounting for the agent's action at every time step. Our approach also enables stability analysis and better control over gradients through time. Taken together, these advantages result in significant improvement over the existing approaches, both in the efficiency and the accuracy of modeling dynamics over extended horizons. We also report promising experimental results in dynamics modeling for the scenarios of both model-based planning and model-free RL.

Abstract: Western countries rely heavily on wheat, and yield prediction is crucial. Time-series deep learning models, such as Long Short Term Memory (LSTM), have already been explored and applied to yield prediction. Existing literature reported that they perform better than traditional Machine Learning (ML) models. However, the existing LSTM cannot handle heterogeneous datasets (a combination of data which varies and remains static with time). In this paper, we propose an efficient deep learning model that can deal with heterogeneous datasets. We developed the system architecture and applied it to the real-world dataset in the digital agriculture area. We showed that it outperforms the existing ML models.

Abstract: Winter wheat is one of the most important crops in the United Kingdom, and crop yield prediction is essential for the nation's food security. Several studies have employed machine learning (ML) techniques to predict crop yield on a county or farm-based level. The main objective of this study is to predict winter wheat crop yield using ML models on multiple heterogeneous datasets, i.e., soil and weather on a zone-based level. Experimental results demonstrated their impact when used alone and in combination. In addition, we employ numerous ML algorithms to emphasize the significance of data quality in any machine-learning strategy.