arXiv subjects

Yang Yu

Publications and source records attributed to Yang Yu.

At least 253 records · Page 14Linked to original sources

arXiv2023

Policy Regularization with Dataset Constraint for Offline Reinforcement Learning

We consider the problem of learning the best possible policy from a fixed dataset, known as offline Reinforcement Learning (RL). A common taxonomy of existing offline RL works is policy regularization, which typically constrains the learned policy by distribution or support of the behavior policy. However, distribution and support constraints are overly conservative since they both force the policy to choose similar actions as the behavior policy when considering particular states. It will limit the learned policy's performance, especially when the behavior policy is sub-optimal. In this paper, we find that regularizing the policy towards the nearest state-action pair can be more effective and thus propose Policy Regularization with Dataset Constraint (PRDC). When updating the policy in a given state, PRDC searches the entire dataset for the nearest state-action sample and then restricts the policy with the action of this sample. Unlike previous works, PRDC can guide the policy with proper behaviors from the dataset, allowing it to choose actions that do not appear in the dataset along with the given state. It is a softer constraint but still keeps enough conservatism from out-of-distribution actions. Empirical evidence and theoretical analysis show that PRDC can alleviate offline RL's fundamentally challenging value overestimation issue with a bounded performance gap. Moreover, on a set of locomotion and navigation tasks, PRDC achieves state-of-the-art performance compared with existing methods. Code is available at https://github.com/LAMDA-RL/PRDC

cs.LG↗

arXiv2023

Transformer-Based Denoising of Mechanical Vibration Signals

Mechanical vibration signal denoising is a pivotal task in various industrial applications, including system health monitoring and failure prediction. This paper introduces a novel deep learning transformer-based architecture specifically tailored for denoising mechanical vibration signals. The model leverages a Multi-Head Attention layer with 8 heads, processing input sequences of length 128, embedded into a 64-dimensional space. The architecture also incorporates Feed-Forward Neural Networks, Layer Normalization, and Residual Connections, resulting in enhanced recognition and extraction of essential features. Through a training process guided by the Mean Squared Error loss function and optimized using the Adam optimizer, the model demonstrates remarkable effectiveness in filtering out noise while preserving critical information related to mechanical vibrations. The specific design and choice of parameters offer a robust method adaptable to the complex nature of mechanical systems, with promising applications in industrial monitoring and maintenance. This work lays the groundwork for future exploration and optimization in the field of mechanical signal analysis and presents a significant step towards advanced and intelligent mechanical system diagnostics.

eess.SY↗

arXiv2023

A Clustering-guided Contrastive Fusion for Multi-view Representation Learning

The past two decades have seen increasingly rapid advances in the field of multi-view representation learning due to it extracting useful information from diverse domains to facilitate the development of multi-view applications. However, the community faces two challenges: i) how to learn robust representations from a large amount of unlabeled data to against noise or incomplete views setting, and ii) how to balance view consistency and complementary for various downstream tasks. To this end, we utilize a deep fusion network to fuse view-specific representations into the view-common representation, extracting high-level semantics for obtaining robust representation. In addition, we employ a clustering task to guide the fusion network to prevent it from leading to trivial solutions. For balancing consistency and complementary, then, we design an asymmetrical contrastive strategy that aligns the view-common representation and each view-specific representation. These modules are incorporated into a unified method known as CLustering-guided cOntrastiVE fusioN (CLOVEN). We quantitatively and qualitatively evaluate the proposed method on five datasets, demonstrating that CLOVEN outperforms 11 competitive multi-view learning methods in clustering and classification. In the incomplete view scenario, our proposed method resists noise interference better than those of our competitors. Furthermore, the visualization analysis shows that CLOVEN can preserve the intrinsic structure of view-specific representation while also improving the compactness of view-commom representation. Our source code will be available soon at https://github.com/guanzhou-ke/cloven.

cs.CV↗

arXiv2023

Disentangling Multi-view Representations Beyond Inductive Bias

Multi-view (or -modality) representation learning aims to understand the relationships between different view representations. Existing methods disentangle multi-view representations into consistent and view-specific representations by introducing strong inductive biases, which can limit their generalization ability. In this paper, we propose a novel multi-view representation disentangling method that aims to go beyond inductive biases, ensuring both interpretability and generalizability of the resulting representations. Our method is based on the observation that discovering multi-view consistency in advance can determine the disentangling information boundary, leading to a decoupled learning objective. We also found that the consistency can be easily extracted by maximizing the transformation invariance and clustering consistency between views. These observations drive us to propose a two-stage framework. In the first stage, we obtain multi-view consistency by training a consistent encoder to produce semantically-consistent representations across views as well as their corresponding pseudo-labels. In the second stage, we disentangle specificity from comprehensive representations by minimizing the upper bound of mutual information between consistent and comprehensive representations. Finally, we reconstruct the original data by concatenating pseudo-labels and view-specific representations. Our experiments on four multi-view datasets demonstrate that our proposed method outperforms 12 comparison methods in terms of clustering and classification performance. The visualization results also show that the extracted consistency and specificity are compact and interpretable. Our code can be found at \url{https://github.com/Guanzhou-Ke/DMRIB}.

cs.CV↗

arXiv2023

Car-Studio: Learning Car Radiance Fields from Single-View and Endless In-the-wild Images

Compositional neural scene graph studies have shown that radiance fields can be an efficient tool in an editable autonomous driving simulator. However, previous studies learned within a sequence of autonomous driving datasets, resulting in unsatisfactory blurring when rotating the car in the simulator. In this letter, we propose a pipeline for learning unconstrained images and building a dataset from processed images. To meet the requirements of the simulator, which demands that the vehicle maintain clarity when the perspective changes and that the contour remains sharp from the background to avoid artifacts when editing, we design a radiation field of the vehicle, a crucial part of the urban scene foreground. Through experiments, we demonstrate that our model achieves competitive performance compared to baselines. Using the datasets built from in-the-wild images, our method gradually presents a controllable appearance editing function. We will release the dataset and code on https://lty2226262.github.io/car-studio/ to facilitate further research in the field.

cs.CV↗

arXiv2023

Distributed online optimization for heterogeneous linear multi-agent systems with coupled constraints

This paper studies a class of distributed online convex optimization problems for heterogeneous linear multi-agent systems. Agents in a network, knowing only their own outputs, need to minimize the time-varying costs through neighboring interaction subject to time-varying coupled inequality constraints. Based on the saddle-point technique, we design a continuous-time distributed controller which is shown to achieve constant regret bound and sublinear fit bound, matching those of the standard centralized online method. We further extend the control law to the event-triggered communication mechanism and show that the constant regret bound and sublinear fit bound are still achieved while reducing the communication frequency. Additionally, we study the situation of communication noise, i.e., the agent's measurement of the relative states of its neighbors is disturbed by a noise. It is shown that, if the noise is not excessive, the regret and fit bounds are unaffected, which indicates the controller's noise-tolerance capability to some extent. Finally, a numerical simulation is provided to support the theoretical conclusions.

math.OC↗

arXiv2023

Uncertainty Estimation by Fisher Information-based Evidential Deep Learning

Uncertainty estimation is a key factor that makes deep learning reliable in practical applications. Recently proposed evidential neural networks explicitly account for different uncertainties by treating the network's outputs as evidence to parameterize the Dirichlet distribution, and achieve impressive performance in uncertainty estimation. However, for high data uncertainty samples but annotated with the one-hot label, the evidence-learning process for those mislabeled classes is over-penalized and remains hindered. To address this problem, we propose a novel method, Fisher Information-based Evidential Deep Learning ($\mathcal{I}$-EDL). In particular, we introduce Fisher Information Matrix (FIM) to measure the informativeness of evidence carried by each sample, according to which we can dynamically reweight the objective loss terms to make the network more focused on the representation learning of uncertain classes. The generalization ability of our network is further improved by optimizing the PAC-Bayesian bound. As demonstrated empirically, our proposed method consistently outperforms traditional EDL-related algorithms in multiple uncertainty estimation tasks, especially in the more challenging few-shot classification settings.

cs.LG↗

arXiv2023

A Unified View of Deep Learning for Reaction and Retrosynthesis Prediction: Current Status and Future Challenges

Reaction and retrosynthesis prediction are fundamental tasks in computational chemistry that have recently garnered attention from both the machine learning and drug discovery communities. Various deep learning approaches have been proposed to tackle these problems, and some have achieved initial success. In this survey, we conduct a comprehensive investigation of advanced deep learning-based models for reaction and retrosynthesis prediction. We summarize the design mechanisms, strengths, and weaknesses of state-of-the-art approaches. Then, we discuss the limitations of current solutions and open challenges in the problem itself. Finally, we present promising directions to facilitate future research. To our knowledge, this paper is the first comprehensive and systematic survey that seeks to provide a unified understanding of reaction and retrosynthesis prediction.

cs.LG↗

arXiv2023

Learning World Models with Identifiable Factorization

Extracting a stable and compact representation of the environment is crucial for efficient reinforcement learning in high-dimensional, noisy, and non-stationary environments. Different categories of information coexist in such environments -- how to effectively extract and disentangle these information remains a challenging problem. In this paper, we propose IFactor, a general framework to model four distinct categories of latent state variables that capture various aspects of information within the RL system, based on their interactions with actions and rewards. Our analysis establishes block-wise identifiability of these latent variables, which not only provides a stable and compact representation but also discloses that all reward-relevant factors are significant for policy learning. We further present a practical approach to learning the world model with identifiable blocks, ensuring the removal of redundants but retaining minimal and sufficient information for policy optimization. Experiments in synthetic worlds demonstrate that our method accurately identifies the ground-truth latent variables, substantiating our theoretical findings. Moreover, experiments in variants of the DeepMind Control Suite and RoboDesk showcase the superior performance of our approach over baselines.

cs.LG↗

arXiv2023

Does Debiasing Inevitably Degrade the Model Performance

Gender bias in language models has attracted sufficient attention because it threatens social justice. However, most of the current debiasing methods degraded the model's performance on other tasks while the degradation mechanism is still mysterious. We propose a theoretical framework explaining the three candidate mechanisms of the language model's gender bias. We use our theoretical framework to explain why the current debiasing methods cause performance degradation. We also discover a pathway through which debiasing will not degrade the model performance. We further develop a causality-detection fine-tuning approach to correct gender bias. The numerical experiment demonstrates that our method is able to lead to double dividends: partially mitigating gender bias while avoiding performance degradation.

cs.CL↗

arXiv2023

NPVForensics: Jointing Non-critical Phonemes and Visemes for Deepfake Detection

Deepfake technologies empowered by deep learning are rapidly evolving, creating new security concerns for society. Existing multimodal detection methods usually capture audio-visual inconsistencies to expose Deepfake videos. More seriously, the advanced Deepfake technology realizes the audio-visual calibration of the critical phoneme-viseme regions, achieving a more realistic tampering effect, which brings new challenges. To address this problem, we propose a novel Deepfake detection method to mine the correlation between Non-critical Phonemes and Visemes, termed NPVForensics. Firstly, we propose the Local Feature Aggregation block with Swin Transformer (LFA-ST) to construct non-critical phoneme-viseme and corresponding facial feature streams effectively. Secondly, we design a loss function for the fine-grained motion of the talking face to measure the evolutionary consistency of non-critical phoneme-viseme. Next, we design a phoneme-viseme awareness module for cross-modal feature fusion and representation alignment, so that the modality gap can be reduced and the intrinsic complementarity of the two modalities can be better explored. Finally, a self-supervised pre-training strategy is leveraged to thoroughly learn the audio-visual correspondences in natural videos. In this manner, our model can be easily adapted to the downstream Deepfake datasets with fine-tuning. Extensive experiments on existing benchmarks demonstrate that the proposed approach outperforms state-of-the-art methods.

cs.CV↗

arXiv2023

Provably Efficient Adversarial Imitation Learning with Unknown Transitions

Imitation learning (IL) has proven to be an effective method for learning good policies from expert demonstrations. Adversarial imitation learning (AIL), a subset of IL methods, is particularly promising, but its theoretical foundation in the presence of unknown transitions has yet to be fully developed. This paper explores the theoretical underpinnings of AIL in this context, where the stochastic and uncertain nature of environment transitions presents a challenge. We examine the expert sample complexity and interaction complexity required to recover good policies. To this end, we establish a framework connecting reward-free exploration and AIL, and propose an algorithm, MB-TAIL, that achieves the minimax optimal expert sample complexity of $\widetilde{O} (H^{3/2} |S|/\varepsilon)$ and interaction complexity of $\widetilde{O} (H^{3} |S|^2 |A|/\varepsilon^2)$. Here, $H$ represents the planning horizon, $|S|$ is the state space size, $|A|$ is the action space size, and $\varepsilon$ is the desired imitation gap. MB-TAIL is the first algorithm to achieve this level of expert sample complexity in the unknown transition setting and improves upon the interaction complexity of the best-known algorithm, OAL, by $O(H)$. Additionally, we demonstrate the generalization ability of MB-TAIL by extending it to the function approximation setting and proving that it can achieve expert sample and interaction complexity independent of $|S|$

cs.LG↗

arXiv2023

Doubly Stochastic Graph-based Non-autoregressive Reaction Prediction

Organic reaction prediction is a critical task in drug discovery. Recently, researchers have achieved non-autoregressive reaction prediction by modeling the redistribution of electrons, resulting in state-of-the-art top-1 accuracy, and enabling parallel sampling. However, the current non-autoregressive decoder does not satisfy two essential rules of electron redistribution modeling simultaneously: the electron-counting rule and the symmetry rule. This violation of the physical constraints of chemical reactions impairs model performance. In this work, we propose a new framework called that combines two doubly stochastic self-attention mappings to obtain electron redistribution predictions that follow both constraints. We further extend our solution to a general multi-head attention mechanism with augmented constraints. To achieve this, we apply Sinkhorn's algorithm to iteratively update self-attention mappings, which imposes doubly conservative constraints as additional informative priors on electron redistribution modeling. We theoretically demonstrate that our can simultaneously satisfy both rules, which the current decoder mechanism cannot do. Empirical results show that our approach consistently improves the predictive performance of non-autoregressive models and does not bring an unbearable additional computational cost.

physics.chem-ph↗

arXiv2023

Learning Physically Realizable Skills for Online Packing of General 3D Shapes

We study the problem of learning online packing skills for irregular 3D shapes, which is arguably the most challenging setting of bin packing problems. The goal is to consecutively move a sequence of 3D objects with arbitrary shapes into a designated container with only partial observations of the object sequence. Meanwhile, we take physical realizability into account, involving physics dynamics and constraints of a placement. The packing policy should understand the 3D geometry of the object to be packed and make effective decisions to accommodate it in the container in a physically realizable way. We propose a Reinforcement Learning (RL) pipeline to learn the policy. The complex irregular geometry and imperfect object placement together lead to huge solution space. Direct training in such space is prohibitively data intensive. We instead propose a theoretically-provable method for candidate action generation to reduce the action space of RL and the learning burden. A parameterized policy is then learned to select the best placement from the candidates. Equipped with an efficient method of asynchronous RL acceleration and a data preparation process of simulation-ready training sequences, a mature packing policy can be trained in a physics-based environment within 48 hours. Through extensive evaluation on a variety of real-life shape datasets and comparisons with state-of-the-art baselines, we demonstrate that our method outperforms the best-performing baseline on all datasets by at least 12.8% in terms of packing utility.

cs.LG↗

arXiv2023

Liouville heat kernel upper bounds at large distances

We show that the Liouville heat kernel decays fast at large distances. In particular, the Liouville semigroup $T_t$ is $C_0$-Feller, where $C_0$ is the space of real-valued continuous functions on $\mathbb C$ vanishing at infinity.

math.PR↗

arXiv2023

A spectral radius condition for a graph to have $(a,b)$-parity factors

Let $a,b$ be two positive integers such that $a \le b$ and $a \equiv b$ (mod $2$). We say that a graph $G$ has an $(a,b)$-parity factor if $G$ has a spanning subgraph $F$ such that $d_{F}(v) \equiv b$ (mod $2$) and $a \le d_{F}(v) \le b$ for all $v \in V (G)$. In this paper, we provide a tight spectral radius condition for a graph to have $(a,b)$-parity factors.

math.CO↗

arXiv2023

Language Model Self-improvement by Reinforcement Learning Contemplation

Large Language Models (LLMs) have exhibited remarkable performance across various natural language processing (NLP) tasks. However, fine-tuning these models often necessitates substantial supervision, which can be expensive and time-consuming to obtain. This paper introduces a novel unsupervised method called LanguageModel Self-Improvement by Reinforcement Learning Contemplation (SIRLC) that improves LLMs without reliance on external labels. Our approach is grounded in the observation that it is simpler for language models to assess text quality than to generate text. Building on this insight, SIRLC assigns LLMs dual roles as both student and teacher. As a student, the LLM generates answers to unlabeled questions, while as a teacher, it evaluates the generated text and assigns scores accordingly. The model parameters are updated using reinforcement learning to maximize the evaluation score. We demonstrate that SIRLC can be applied to various NLP tasks, such as reasoning problems, text generation, and machine translation. Our experiments show that SIRLC effectively improves LLM performance without external supervision, resulting in a 5.6% increase in answering accuracy for reasoning tasks and a rise in BERTScore from 0.82 to 0.86 for translation tasks. Furthermore, SIRLC can be applied to models of different sizes, showcasing its broad applicability.

cs.CL↗

arXiv2023

Compact Lattice Gadget and Its Applications to Hash-and-Sign Signatures

This work aims to improve the practicality of gadget-based cryptosystems, with a focus on hash-and-sign signatures. To this end, we develop a compact gadget framework in which the used gadget is a square matrix instead of the short and fat one used in previous constructions. To work with this compact gadget, we devise a specialized gadget sampler, called semi-random sampler, to compute the approximate preimage. It first deterministically computes the error and then randomly samples the preimage. We show that for uniformly random targets, the preimage and error distributions are simulatable without knowing the trapdoor. This ensures the security of the signature applications. Compared to the Gaussian-distributed errors in previous algorithms, the deterministic errors have a smaller size, which lead to a substantial gain in security and enables a practically working instantiation. As the applications, we present two practically efficient gadget-based signature schemes based on NTRU and Ring-LWE respectively. The NTRU-based scheme offers comparable efficiency to Falcon and Mitaka and a simple implementation without the need of generating the NTRU trapdoor. The LWE-based scheme also achieves a desirable overall performance. It not only greatly outperforms the state-of-the-art LWE-based hash-and-sign signatures, but also has an even smaller size than the LWE-based Fiat-Shamir signature scheme Dilithium. These results fill the long-term gap in practical gadget-based signatures.

cs.CR↗