检索结果-内蒙古大学图书馆

Self-Imitation Guided High-Efficient Goal-Conditioned Reinforcement Learning

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Li, Yao Wang, YuHui Tan, XiaoYang College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics MIIT Key Laboratory of Pattern Analysis and Machine Intelligence China Saudi Arabia

Goal-conditioned reinforcement learning (GCRL) aims to manipulate robots to reach desired goals, which poses a major challenge based on the varying configurations and objectives in different tasks. The challenges may increase the proportion of meaningless steps when reaching the desired goal, negatively affecting policy performance. In this paper, we propose a goal-conditioned reinforcement learning method guided by self-imitation by using past good decisions for efficient policy training. The past good decisions are derived by maximizing the log-likelihood of the visited trajectories as self-imitated learning, and maximizing the goal-conditioned action values for reaching the goals with fewer meaningless steps. To leverage the advantage of past good decisions, a selfimitation learning operator is proposed to use past decisions for efficient action value learning. Theoretical analysis demonstrates that our method improves monotonically based on the self-imitation guided training mechanism. Experimental results show that our method achieves the most stable performance compared to the baselines in robotic control tasks. © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

PointSmile: Point Self-supervised Learning via Curriculum Mutual Information

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Xin Wei, Mingqiang Chen, Songcan The School of Computer Science and Technology Nanjing University of Aeronautics and Astronautics China The MIIT Key Laboratory of Pattern Analysis and Machine Intelligence Nanjing China

Self-supervised learning is attracting wide attention in point cloud processing. However, it is still not well-solved to gain discriminative and transferable features of point clouds for efficient training on downstream tasks, due to their natural sparsity and irregularity. We propose PointSmile, a reconstruction-free self-supervised learning paradigm by maximizing curriculum mutual information (CMI) across the replicas of point cloud objects. From the perspective of how-and-what-to-learn, PointSmile is designed to imitate human curriculum learning, i.e., starting with an easy curriculum and gradually increasing the difficulty of that curriculum. To solve "how-to-learn", we introduce curriculum data augmentation (CDA) of point clouds. CDA encourages PointSmile to learn from easy samples to hard ones, such that the latent space can be dynamically affected to create better embeddings. To solve "what-to-learn", we propose to maximize both feature- and class-wise CMI, for better extracting discriminative features of point clouds. Unlike most of existing methods, PointSmile does not require a pretext task, nor does it require cross-modal data to yield rich latent representations. We demonstrate the effectiveness and robustness of PointSmile in downstream tasks including object classification and segmentation. Extensive results show that our PointSmile outperforms existing self-supervised methods, and compares favorably with popular fully-supervised methods on various standard architectures. Copyright © 2023, The Authors. All rights reserved.

关键词： Supervised learning

A centroid auto-fused hierarchical fuzzy c-means clustering

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Lin, Yunxia Chen, Songcan College of Computer Science and Technology College of Artificial Intelligence Nanjing University of Aeronautics and Astronautics Nanjing211106 China MIIT Key Laboratory of Pattern Analysis and Machine Intelligence

Like k-means and Gaussian Mixture Model (GMM), fuzzy c-means (FCM) with soft partition has also become a popular clustering algorithm and still is extensively studied. However, these algorithms and their variants still suffer from some difficulties such as determination of the optimal number of clusters which is a key factor for clustering quality. A common approach for overcoming this difficulty is to use the trial-and-validation strategy, i.e., traversing every integer from large number like √n to 2 until finding the optimal number corresponding to the peak value of some cluster validity index. But it is scarcely possible to naturally construct an adaptively agglomerative hierarchical cluster structure as using the trial- and-validation strategy. Even possible, existing different validity indices also lead to different number of clusters. To effectively mitigate the problems while motivated by convex clustering, in this paper we present a Centroid Auto-Fused Hierarchical Fuzzy c-means method (CAF-HFCM) whose optimization procedure can automatically agglomerate to form a cluster hierarchy, more importantly, yielding an optimal number of clusters without resorting to any validity index. Although a recently-proposed robust-learning fuzzy c-means (RL-FCM) can also automatically obtain the best number of clusters without the help of any validity index, so-involved 3 hyper-parameters need to adjust expensively, conversely, our CAF-HFCM involves just 1 hyper-parameter which makes the corresponding adjustment is relatively easier and more operational. Further, as an additional benefit from our optimization objective, the CAF-HFCM effectively reduces the sensitivity to the initialization of clustering performance. Moreover, our proposed CAF-HFCM method is able to be straightforwardly extended to various variants of FCM. Finally, extensive experiments on both synthetic and real data sets demonstrate the effectiveness and efficiency of CAF-HFCM. Copyright © 2020, The Authors

关键词： Gaussian distribution

A Simplified Student Network with Multi-teacher Feature Fusion for Industrial Defect Detection 1

学校读者我要写书评

暂无评论

7th Asian Conference on pattern Recognition, ACPR 2023

作者： Pei, Mingjing Liu, Ningzhong College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China School of Electronics and Information Engineering West Anhui University Lu’an China MIIT Key Laboratory of Pattern Analysis and Machine Intelligence Nanjing China Collaborative Innovation Center of Novel Software Technology and Industrialization Nanjing China

ISBN: (数字)9783031476341

ISBN: (纸本)9783031476334

Improved industrial defect detection is deemed critical for ensuring high-quality manufacturing processes. Despite the effectiveness of knowledge distillation in detecting defects, there are still challenges in extracting useful features and designing a better student network. In this study, a new framework based on knowledge distillation is proposed to address these issues. To avoid overfitting and balance anomaly detection between image and pixel levels, a simplified student decoding network is designed. To extract more diverse features, a multi-teacher network is used in the teacher network. Simple addition and concatenation operations for feature maps were not sufficient, the method fuses features iteratively using attention mechanisms utilized for the multi-teacher networks. The approach is evaluated on two benchmark datasets for industrial defect detection, Mvtec and BTAD, and significantly improved performance is achieved compared to the other methods. An average accuracy of 99.22% and 97.46% for image-level and pixel-level ROC-AUC, respectively, is achieved on the Mvtec dataset, and 93.47% and 96.3% on the BTAD dataset. The proposed framework shows effectiveness in detecting defects in industrial settings, as demonstrated by the results. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Distillation

A concise yet effective model for non-aligned incomplete multi-view and missing multi-label learning

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Li, Xiang Chen, Songcan College of Computer Science and Technology College of Artificial Intelligence Nanjing University of Aeronautics and Astronautics Nanjing211106 China MIIT Key Laboratory of Pattern Analysis and Machine Intelligence

—In reality, learning from multi-view multi-label data inevitably confronts three challenges: missing labels, incomplete views, and non-aligned views. Existing methods mainly concern the first two and commonly need multiple assumptions to attack them, making even state-of-the-arts involve at least two explicit hyper-parameters such that model selection is quite difficult. More toughly, they will fail in handling the third challenge, let alone addressing the three jointly. In this paper, we aim at meeting these under the least assumption by building a concise yet effective model with just one hyper-parameter. To ease insufficiency of available labels, we exploit not only the consensus of multiple views but also the global and local structures hidden among multiple labels. Specifically, we introduce an indicator matrix to tackle the first two challenges in a regression form while aligning the same individual labels and all labels of different views in a common label space to battle the third challenge. In aligning, we characterize the global and local structures of multiple labels to be high-rank and low-rank, respectively. Subsequently, an efficient algorithm with linear time complexity in the number of samples is established. Finally, even without view-alignment, our method substantially outperforms state-of-the-arts with view-alignment on five real datasets. Copyright © 2020, The Authors. All rights reserved.

关键词： machine learning

Pushing One Pair of Labels Apart Each Time in Multi-Label Learning: From Single Positive to Full Labels

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Xiang Wang, Xinrui Chen, Songcan College of Computer Science and Technology College of Artificial Intelligence Nanjing University of Aeronautics and Astronautics Nanjing211106 China MIIT Key Laboratory of Pattern Analysis and Machine Intelligence

In Multi-Label Learning (MLL), it is extremely challenging to accurately annotate every appearing object due to expensive costs and limited knowledge. When facing such a challenge, a more practical and cheaper alternative should be Single Positive Multi-Label Learning (SPMLL), where only one positive label needs to be provided per sample. Existing SPMLL methods usually assume unknown labels as negatives, which inevitably introduces false negatives as noisy labels. More seriously, Binary Cross Entropy (BCE) loss is often used for training, which is notoriously not robust to noisy labels. To mitigate this issue, we customize an objective function for SPMLL by pushing only one pair of labels apart each time to prevent the domination of negative labels, which is the main culprit of fitting noisy labels in SPMLL. To further combat such noisy labels, we explore the high-rankness of label matrix, which can also push apart different labels. By directly extending from SPMLL to MLL with full labels, a unified loss applicable to both settings is derived. Experiments on real datasets demonstrate that the proposed loss not only performs more robustly to noisy labels for SPMLL but also works well for full labels. Besides, we empirically discover that high-rankness can mitigate the dramatic performance drop in SPMLL. Most surprisingly, even without any regularization or fine-tuned label correction, only adopting our loss defeats state-of-the-art SPMLL methods on CUB, a dataset that severely lacks labels. Copyright © 2023, The Authors. All rights reserved.

关键词： machine learning

Near-Optimal Decentralized Momentum Method for Nonconvex-PL Minimax Problems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Huang, Feihu Chen, Songcan College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China Miit Key Laboratory of Pattern Analysis and Machine Intelligence Nanjing China

Minimax optimization plays an important role in many machine learning tasks such as generative adversarial networks (GANs) and adversarial training. Although recently a wide variety of optimization methods have been proposed to solve the minimax problems, most of them ignore the distributed setting where the data is distributed on multiple workers. Meanwhile, the existing decentralized minimax optimization methods rely on the strictly assumptions such as (strongly) concavity and variational inequality conditions. In the paper, thus, we propose an efficient decentralized momentum-based gradient descent ascent (DM-GDA) method for the distributed nonconvex-PL minimax optimization, which is nonconvex in primal variable and is nonconcave in dual variable and satisfies the Polyak-Lojasiewicz (PL) condition. In particular, our DM-GDA method simultaneously uses the momentum-based techniques to update variables and estimate the stochastic gradients. Moreover, we provide a solid convergence analysis for our DM-GDA method, and prove that it obtains a near-optimal gradient complexity of O(ϵ-3) for finding an ϵ-stationary solution of the nonconvex-PL stochastic minimax problems, which reaches the lower bound of nonconvex stochastic optimization. To the best of our knowledge, we first study the decentralized algorithm for Nonconvex-PL stochastic minimax optimization over a network. © 2023, CC BY-NC-ND.

关键词： Momentum

Rectified euler k-means and beyond

学校读者我要写书评

暂无评论

arXiv 2021年

Euler k-means (EulerK) first maps data onto the unit hyper-sphere surface of equi-dimensional space via a complex mapping which induces the robust Euler kernel and next employs the popular k-means. Consequently, besides enjoying the virtues of k-means such as simplicity and scalability to large data sets, EulerK is also robust to noises and outliers. Although so, the centroids captured by EulerK deviate from the unit hyper-sphere surface and thus in strict distributional sense, actually are outliers. This weird phenomenon also occurs in some generic kernel clustering methods. Intuitively, using such outlier-like centroids should not be quite reasonable but it is still seldom attended. To eliminate the deviation, we propose two Rectified Euler k-means methods, i.e., REK1 and REK2, which retain the merits of EulerK while acquire real centroids residing on the mapped space to better characterize the data structures. Specifically, REK1 rectifies EulerK by imposing the constraint on the centroids while REK2 views each centroid as the mapped image from a pre-image in the original space and optimizes these pre-images in Euler kernel induced space. Undoubtedly, our proposed REKs can methodologically be extended to solve problems of such a category. Finally, the experiments validate the effectiveness of REK1 and REK2. © 2021, CC BY.

关键词： Statistics

Transport based Graph Kernels

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Ma, Kai Wan, Peng Zhang, Daoqiang Department of Computer Science and Technology MIIT Key Laboratory of Pattern Analysis and Machine Intelligence Nanjing University of Aeronautics and Astronautics Jiangsu Province210016 China

Graph kernel is a powerful tool measuring the similarity between graphs. Most of the existing graph kernels focused on node labels or attributes and ignored graph hierarchical structure information. In order to effectively utilize graph hierarchical structure information, we propose pyramid graph kernel based on optimal transport (OT). Each graph is embedded into hierarchical structures of the pyramid. Then, the OT distance is utilized to measure the similarity between graphs in hierarchical structures. We also utilize the OT distance to measure the similarity between subgraphs and propose subgraph kernel based on OT. The positive semidefinite (p.s.d) of graph kernels based on optimal transport distance is not necessarily possible. We further propose regularized graph kernel based on OT where we add the kernel regularization to the original optimal transport distance to obtain p.s.d kernel matrix. We evaluate the proposed graph kernels on several benchmark classification tasks and compare their performance with the existing state-of-the-art graph kernels. In most cases, our proposed graph kernel algorithms outperform the competing methods. Copyright © 2020, The Authors. All rights reserved.

关键词： Benchmarking