检索结果-内蒙古大学图书馆

Fast One-Pass Approximation for Matrix-Based Rényi's Entropy

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Gong, Tieliang Dong, Yuxin Gao, Zeyu Chen, Hong Zhang, Weizhan Shaanxi Xi’an710049 China Shaanxi Provincial Key Laboratory of Big Data Knowledge Engineering Shaanxi Xi’an710049 China Department of Oncology University of Cambridge CambridgeCB2 1TN United Kingdom Department of Mathematics and Statistics Huazhong Agriculture University Hubei Wuhan430070 China

This paper is concerned with improving the computational efficiency of the matrix-based Rényi's α-order entropy, which enables directly being accessed from the given data without knowing the probability density function of the underlying data. However, leveraging either trace estimation or the eigenvalue-based methods will cost O(n3) in time complexity, where n is the number of observed samples. The intensive computational demands impede its utility in modern large-scale machine learning tasks. Recently, the Hutch++ based entropy approximation was developed, by converting the original entropy estimation problem into stochastic trace estimation. It reduces the time complexity to O(n2s) (s © 2024, The Authors. All rights reserved.

关键词： Distillation

Block-Structured Optimization for Anomalous Pattern Detection in Interdependent Networks

学校读者我要写书评

暂无评论

Block-Structured Optimization for Anomalous Pattern Detectio...

IEEE International Conference on data Mining (ICDM)

作者： Fei Jie Chunpai Wang Feng Chen Lei Li Xindong Wu Key Laboratory of Knowledge Engineering with Big Data (Hefei University of Technology) Ministry of Education Hefei China School of Computer Science and Information Engineering Hefei University of Technology Hefei China Department of Computer Science University at Albany – SUNY Albany NY USA Erik Jonsson School of Engineering & Computer Science The University of Texas at Dallas Dallas TX USA Mininglamp Academy of Sciences Mininglamp Technology Beijing China Institute of Big Knowledge Science Hefei University of Technology Hefei China

We propose a generalized optimization framework for detecting anomalous patterns (subgraphs that are interesting or unexpected) in interdependent networks, such as multi-layer networks, temporal networks, networks of networks, and many others. We frame the problem as a non-convex optimization that has a general nonlinear score function and a set of block-structured and non-convex constraints. We develop an effective, efficient, and parallelizable projection-based algorithm, namely Graph Block-structured Gradient Projection (GBGP), to solve the problem. It is proved that our algorithm 1) runs in nearly-linear time on the network size, and 2) enjoys a theoretical approximation guarantee. Moreover, we demonstrate how our framework can be applied to two very practical applications, and we conduct comprehensive experiments to show the effectiveness and efficiency of our proposed algorithm.

关键词：

Block-Structured Optimization for Subgraph Detection in Interdependent Networks

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jie, Fei Wang, Chunpai Chen, Feng Li, Lei Wu, Xindong Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology Ministry of Education Hefei China School of Computer Science and Information Engineering Hefei University of Technology Hefei China Department of Computer Science University at Albany – SUNY AlbanyNY United States Erik Jonsson School of Engineering & Computer Science The University of Texas at Dallas DallasTX United States Mininglamp Academy of Sciences Mininglamp Technologies Beijing China Institute of Big Knowledge Science Hefei University of Technology Hefei China

We propose a generalized framework for block-structured nonconvex optimization, which can be applied to structured subgraph detection in interdependent networks, such as multi-layer networks, temporal networks, networks of networks, and many others. Specifically, we design an effective, efficient, and parallelizable projection algorithm, namely Graph Block-structured Gradient Projection (GBGP), to optimize a general non-linear function subject to graph-structured constraints. We prove that our algorithm: 1) runs in nearly-linear time on the network size;2) enjoys a theoretical approximation guarantee. Moreover, we demonstrate how our framework can be applied to two very practical applications and conduct comprehensive experiments to show the effectiveness and efficiency of our proposed algorithm. Copyright © 2022, The Authors. All rights reserved.

关键词： Functions

An Empirical Study on Information Extraction using Large Language Models

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Han, Ridong Yang, Chaohao Peng, Tao Tiwari, Prayag Wan, Xiang Liu, Lu Wang, Benyou College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University China College of Software Jilin University China Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen China School of Data Science The Chinese University of Hong Kong Shenzhen China School of Information Technology Halmstad University Sweden

Human-like large language models (LLMs), especially the most powerful and popular ones in OpenAI’s GPT family, have proven to be very helpful for many natural language processing (NLP) related tasks. Therefore, various attempts have been made to apply LLMs to information extraction (IE), which is a fundamental NLP task that involves extracting information from unstructured plain text. To demonstrate the latest representative progress in LLMs’ information extraction ability, we assess the information extraction ability of GPT-4 (the latest version of GPT at the time of writing this paper) from four perspectives: Performance, Evaluation Criteria, Robustness, and Error Types. Our results suggest a visible performance gap between GPT-4 and state-of-the-art (SOTA) IE methods. To alleviate this problem, considering the LLMs’ human-like characteristics, we propose and analyze the effects of a series of simple prompt-based methods, which can be generalized to other LLMs and NLP tasks. Rich experiments show our methods’ effectiveness and some of their remaining issues in improving GPT-4’s information extraction ability. Copyright © 2023, The Authors. All rights reserved.

关键词： Information retrieval

Strategy intervention for the evolution of fairness

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Zhang, Yanling Fu, Feng Key Laboratory of Knowledge Automation for Industrial Processes of Ministry of Education School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing China Department of Mathematics Dartmouth College HanoverNH United States Department of Biomedical Data Science Geisel School of Medicine Dartmouth College HanoverNH United States

Masses of experiments have shown individual preference for fairness which seems irrational. The reason behind it remains a focus for research. The effect of spite (individuals are only concerned with their own relative standing) on the evolution of fairness has attracted increasing attention from experiments, but only has been implicitly studied in one evolutionary model. The model did not involve high-offer rejections, which have been found in the form of non-monotonic rejections (rejecting offers that are too high or too low) in experiments. Here, we introduce a high offer and a non-monotonic rejection in structured populations of finite size, and use strategy intervention to explicitly study how spite influences the evolution of fairness: five strategies are in sequence added into the competition of a fair strategy and a selfish strategy. We find that spite promotes fairness, altruism inhibits fairness, and the non-monotonic rejection can cause fairness to overcome selfishness, which cannot happen without high-offer rejections. Particularly for the group-structured population with seven discrete strategies, we analytically study the effect of population size, mutation, and migration on fairness, selfishness, altruism, and spite. A larger population size cannot change the dominance of fairness, but it promotes altruism and inhibits selfishness and spite. Intermediate mutation maximizes selfishness and fairness, and minimizes spite;intermediate mutation maximizes altruism for intermediate migration and minimizes altruism otherwise. The existence of migration inhibits selfishness and fairness, and promotes altruism;sufficient migration promotes spite. Our study may provide important insights into the evolutionary origin of fairness. Copyright © 2017, The Authors. All rights reserved.

关键词： Population statistics

Intervention Prediction for Patients with Pressure Injury Using Random Forest

学校读者我要写书评

暂无评论

Intervention Prediction for Patients with Pressure Injury Us...

IEEE International Conference on Big knowledge (ICBK)

作者： Liuqi Jin Yan Pan Jiaoyun Yang Lin Han Lin Lv Miki Raviv Ning An Key Laboratory of Knowledge Engineering with Big Data of the Ministry of Education School of Computer Science and Information Engineering Hefei University of Technology Hefei China Evidence-Based Nursing Center School of Nursing Lanzhou University Lanzhou China Gansu Provincal Hospital Wound and Ostomy Care Center Lanzhou China Vitalerter LTD Ha-Yarden Airport City Israel

ISBN: (纸本)9781665438599

Pressure injury (PI) is one of the major causes of short-term death. Early intervention for patients at risk plays an essential role in PI. However, many nurses may ignore risks. This paper aims to establish a model to predict interventions according to the patient's physical signs, which can help nurses develop care plans. We used data from 1,483 patients with 25 characteristics and 17 interventions. We use the Random Forest and Particle Swarm Optimization (PSO) to optimize model parameters. Then we compared it with KNN, SVM, and Decision Tree. The 10-fold cross-validation result showed that the Random Forest has better accuracy than other methods, with an f1 score of 0.84. This finding proved the feasibility of using machine learning to help formulate care plans according to the classification of index prediction results. Our model shows that hemoglobin, Braden PI score, and age are the three most influential risk factors.

关键词： Support vector machines Conferences Forestry Predictive models Indexes Decision trees Particle swarm optimization

Hierarchical GAN-Tree and Bi-Directional Capsules for multi-label image classification

学校读者我要写书评

暂无评论

knowledge-Based Systems 2022年 238卷

作者： Wang, Boyan Hu, Xuegang Zhang, Chenwei Li, Peipei Yu, Philip S. Hefei University of Technology Anhui Province Hefei China Key Laboratory of Knowledge Engineering with Big Data (Hefei University of Technology) Anhui Province Hefei China Anhui Province Key Laboratory of Industry Safety and Emergency Technology Anhui Province Hefei China *** Inc. Seattle WA United States University of Illinois at Chicago Chicago IL United States

Compared with the flat multi-label image classification, the hierarchical structure reserves a richer source of structural information to represent complicated relationships between labels in the real world. However, existing multi-label image classification methods focus on the accuracy of label prediction, ignoring the structural information embedded in the hierarchical label space. Furthermore, they hardly form the relevant visual feature space corresponding to the hierarchical label structure. In this paper, we propose a novel hierarchical framework based on the feature and label structural information named Hierarchical GAN-Tree and Bi-Directional Capsules (HGT&BC) to address these problems. We conduct Hierarchical GAN-Tree for feature space representation and Hierarchical Bi-Directional Capsules for label space classification, respectively. Hierarchical GAN-Tree generates hierarchical feature space using the unsupervised divisive clustering pattern according to the hierarchical structure, alleviating the mode-collapse of generators and the overfitting manifestation of conventional GANs. Hierarchical Bi-Directional Capsules utilize the hierarchical label structure in iterations of top-down and bottom-up processes: the top-down process integrates hierarchical relationships into the probability computation to enhance partial hierarchical relationships;the bottom-up process modifies the dynamic routing mechanism between capsules to represent semantic objects for the comprehensive global hierarchical classifiers. Owing to the two components, HGT&BC successfully expresses the hierarchical relationships in both feature and label space and improves the performance of multi-label image classification. Extensive experimental results on four benchmark datasets demonstrate the effectiveness and efficiency of our hierarchical framework in practice. © 2021 Elsevier B.V.

关键词： Capsule Neural Network Generative Adversarial Network Hierarchical learning Multi-label image classification

ReconBoost: Boosting Can Achieve Modality Reconcilement

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hua, Cong Xu, Qianqian Bao, Shilong Yang, Zhiyong Huang, Qingming Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China Institute of Information Engineering Chinese Academy of Sciences Beijing China School of Cyber Security University of Chinese Academy of Sciences Beijing China Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of Sciences Beijing China

This paper explores a novel multi-modal alternating learning paradigm pursuing a reconciliation between the exploitation of uni-modal features and the exploration of cross-modal interactions. This is motivated by the fact that current paradigms of multi-modal learning tend to explore multi-modal features simultaneously. The resulting gradient prohibits further exploitation of the features in the weak modality, leading to modality competition, where the dominant modality overpowers the learning process. To address this issue, we study the modality-alternating learning paradigm to achieve reconcilement. Specifically, we propose a new method called ReconBoost to update a fixed modality each time. Herein, the learning objective is dynamically adjusted with a reconcilement regularization against competition with the historical models. By choosing a KL-based reconcilement, we show that the proposed method resembles Friedman's Gradient-Boosting (GB) algorithm, where the updated learner can correct errors made by others and help enhance the overall performance. The major difference with the classic GB is that we only preserve the newest model for each modality to avoid overfitting caused by ensembling strong learners. Furthermore, we propose a memory consolidation scheme and a global rectification scheme to make this strategy more effective. Experiments over six multi-modal benchmarks speak to the efficacy of the method. We release the code at https://***/huacong/ReconBoost. Copyright © 2024, The Authors. All rights reserved.

关键词： Machine learning

data-Driven Single Image Deraining: A Comprehensive Review and New Perspectives

学校读者我要写书评

暂无评论

TechRxiv

TechRxiv 2021年

作者： Zhang, Zhao Wei, Yanyan Zhang, Haijun Yang, Yi Yan, Shuicheng Wang, Meng School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China Key Laboratory of Knowledge Engineering with Big Data Ministry of Education Hefei University of Technology Hefei230009 China Shenzhen China Centre for Artificial Intelligence University of Technology Sydney SydneyNSW Australia Department of Electrical and Computer Engineering National University of Singapore Singapore Hefei University of Technology China

Single Image Deraining task aims at recovering the rain-free background from an image degraded by rain streaks and rain accumulation. For the powerful fitting ability of deep neural networks and massive training data, data-driven deep SID methods obtained significant improvement over traditional ones. Current SID methods usually focus on improving the deraining performance by proposing different kinds of deraining networks, while neglecting the interpretation of the solving process. As a result, the generalization ability may still be limited in real-world scenarios, and the deraining results also cannot effectively improve the performance of subsequent high-level tasks (e.g., object detection). To explore these issues, we in this paper re-examine the three important factors (i.e., data, rain model and network architecture) for the SID problem, and specifically analyze them by proposing new and more reasonable criteria (i.e., general vs. specific, synthetical vs. mathematical, black-box vs. white-box). We also study the relationship of the three factors from a new perspective of data, and reveal two different solving paradigms (explicit vs. implicit) for the SID task. We further discuss the current mainstream data-driven SID methods from five aspects, i.e., training strategy, network pipeline, domain knowledge, data preprocessing, and objective function, and some useful conclusions are summarized by statistics. Besides, we profoundly studied one of the three factors, i.e., data, and measured the performance of current methods on different datasets through extensive experiments to reveal the effectiveness of SID data. Finally, with the comprehensive review and in-depth analysis, we draw some valuable conclusions and suggestions for future research. © 2021, CC BY.

关键词： Object detection