检索结果-内蒙古大学图书馆

Accessing the Suffix Array via - 1 -Forest

学校读者我要写书评

暂无评论

Accessing the Suffix Array via - 1 -Forest

29th International Symposium on String Processing and Information Retrieval, SPIRE 2022

作者： Boucher, Christina Köppl, Dominik Perera, Herman Rossi, Massimiliano Department of Computer and Information Science and Engineering Herbert Wertheim College of Engineering University of Florida GainesvilleFL United States Tokyo Medical and Dental University M &D Data Science Center Tokyo Japan

ISBN: (纸本)9783031206429

Kärkkainen et al. (CPM, 2009) defined the concept of that later became key to the construction of the r-index. Given a string S[1.n], its suffix array SA and its inverse suffix array ISA, we define as the permutation of { 1, …, n} such that (i) = SA [ ISA [ i] - 1 ] if ISA [ i] > 1, and (i) = SA [ n] otherwise. Gagie et al. (JACM, 2020) showed that it is possible to store O(r) words such that the permutations and - 1 are evaluated in O(log log w(n/ r) ) -time, which was improved to O(1 ) -time by Nishimoto and Tabei (ICALP, 2021). In this paper, we introduce the concept of - 1 -forest, which is a data structure using sampled SA values to speed up random access to SA. We implemented our approach and compared its performance with respect to the r-index. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Forestry

AuRoRA: A One-for-all Platform for Augmented Reasoning and Refining with Task-Adaptive Chain-of-Thought Prompting 30

学校读者我要写书评

暂无评论

AuRoRA: A One-for-all Platform for Augmented Reasoning and R...

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Zou, Anni Zhang, Zhuosheng Zhao, Hai Li, Xiaoshan Bian, Minjie Department of Computer Science and Engineering Shanghai Jiao Tong University China School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University China Shanghai Technology Innovation Department Shanghai Data Group Co. Ltd. China Shanghai Data Group Co. Ltd. China

ISBN: (纸本)9782493814104

Large language models (LLMs) empowered by chain-of-thought (CoT) prompting have yielded remarkable prowess in reasoning tasks. Nevertheless, current methods predominantly lean on handcrafted or task-specific demonstrations, lack reliable knowledge basis and thus struggle for trustworthy responses in an automated pattern. While recent works endeavor to improve upon one certain aspect, they ignore the importance and necessity of establishing an integrated and interpretable reasoning system. To address these drawbacks and provide a universal solution, we propose AuRoRA: a one-for-all platform for augmented reasoning and refining based on CoT prompting that excels in adaptability, reliability, integrity, and interpretability. The system exhibits superior performances across six reasoning tasks and offers real-time visual analysis, which has pivotal academic and application value in the era of LLMs. The AuRoRA platform is available at https://***/spaces/Anni123/AuRoRA. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Refining

Disentanglement with Factor Quantized Variational Autoencoders

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Baykal, Gulcin Kandemir, Melih Unal, Gozde Department of Mathematics and Computer Science University of Southern Denmark Odense Denmark Department of Computer Engineering Istanbul Technical University Istanbul Turkey Department of AI and Data Engineering Istanbul Technical University Istanbul Turkey

Disentangled representation learning aims to represent the underlying generative factors of a dataset in a latent representation independently of one another. In our work, we propose a discrete variational autoencoder (VAE) based model where the ground truth information about the generative factors are not provided to the model. We demonstrate the advantages of learning discrete representations over learning continuous representations in facilitating disentanglement. Furthermore, we propose incorporating an inductive bias into the model to further enhance disentanglement. Precisely, we propose scalar quantization of the latent variables in a latent representation with scalar values from a global codebook, and we add a total correlation term to the optimization as an inductive bias. Our method called FactorQVAE combines optimization based disentanglement approaches with discrete representation learning, and it outperforms the former disentanglement methods in terms of two disentanglement metrics (DCI and InfoMEC) while improving the reconstruction performance. Our code can be found at https://***/ituvisionlab/FactorQVAE. Copyright © 2024, The Authors. All rights reserved.

关键词： Vector quantization

Partial Identification of Causal Effects Using Proxy Variables

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Ghassami, AmirEmad Shpitser, Ilya Tchetgen, Eric Tchetgen Department of Mathematics and Statistics Boston University United States Department of Computer Science Johns Hopkins University United States Department of Statistics and Data Science University of Pennsylvania United States

Proximal causal inference is a recently proposed framework for evaluating the causal effect of a treatment on an outcome variable in the presence of unmeasured confounding (Miao et al., 2018a;Tchetgen Tchetgen et al., 2020). For nonparametric point identification of causal effects, the framework leverages a pair of so-called treatment and outcome confounding proxy variables, in order to identify a bridge function that matches the dependence of potential outcomes or treatment variables on the hidden factors to corresponding functions of observed proxies. Unique identification of a causal effect via a bridge function crucially requires that proxies are sufficiently relevant for hidden factors, a requirement that has previously been formalized as a completeness condition. However, completeness is well-known not to be empirically testable, and although a bridge function may be well-defined in a given setting, lack of completeness, sometimes manifested by availability of a single type of proxy, may severely limit prospects for identification of a bridge function and thus a causal effect;therefore, potentially restricting the application of the proximal causal framework. In this paper, we propose partial identification methods that do not require completeness and obviate the need for identification of a bridge function. That is, we establish that proxies of unobserved confounders can be leveraged to obtain bounds on the causal effect of the treatment on the outcome even if available information does not suffice to identify either a bridge function or a corresponding causal effect of interest. Our bounds are non-smooth functionals of the underlying distribution. As a consequence, in the context of inference, we initially employ the LogSumExp approximation, which represents a smooth approximation of our bounds. Subsequently, we leverage bootstrap confidence intervals on the approximated bounds. We further establish analogous partial identification results in related setting

关键词： Statistical methods

DPMAC: Differentially Private Communication for Cooperative Multi-Agent Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zhao, Canzhe Ze, Yanjie Dong, Jing Wang, Baoxiang Li, Shuai John Hopcroft Center for Computer Science Shanghai Jiao Tong University China Department of Computer Science and Engineering Shanghai Jiao Tong University China School of Data Science The Chinese University of Hong Kong Shenzhen China

Communication lays the foundation for cooperation in human society and in multi-agent reinforcement learning (MARL). Humans also desire to maintain their privacy when communicating with others, yet such privacy concern has not been considered in existing works in MARL. To this end, we propose the differentially private multi-agent communication (DPMAC) algorithm, which protects the sensitive information of individual agents by equipping each agent with a local message sender with rigorous (ϵ, δ)-differential privacy (DP) guarantee. In contrast to directly perturbing the messages with predefined DP noise as commonly done in privacy-preserving scenarios, we adopt a stochastic message sender for each agent respectively and incorporate the DP requirement into the sender, which automatically adjusts the learned message distribution to alleviate the instability caused by DP noise. Further, we prove the existence of a Nash equilibrium in cooperative MARL with privacy-preserving communication, which suggests that this problem is game-theoretically learnable. Extensive experiments demonstrate a clear advantage of DPMAC over baseline methods in privacy-preserving scenarios. Copyright © 2023, The Authors. All rights reserved.

关键词： Stochastic systems

Machine Learning based Enhanced Autonomous Driving for Autonomous Vehicles 6

学校读者我要写书评

暂无评论

Machine Learning based Enhanced Autonomous Driving for Auton...

6th International Conference on Inventive Computation Technologies, ICICT 2023

作者： Lavanya, M.C. Akshatha, M. Radhakrishnan, P. Bai, Kshatriya Vinaya Sree Alakshmi, S. Vijay Ranga, Jarabala Dept of CSE Vidyavardhaka College of Engineering Mysuru India Dept. of Electronics& communication Engineering Tagore Engineering College India Artificial Intelligence & Data Science Lakireddy Bali Reddy College of Engineering India Department of Computer Science and Engineering Park College of Engineering and Technology Tamilnadu Coimbatore641659 India Department of EEE Ramachandra college of engineering Eluru AP India

ISBN: (纸本)9798350398496

In particular, the emphasis of our study is on the most important difficulties associated with finding data-driven representations of nonlinear systems that allow for linear analysis is a key step towards deriving dynamics from data. Many issues plaguing modern dynamical systems are explored and the innovative methods being created in data science and machine learning to address them. Nonlinear dynamics, on the one hand, and unknown or only partially understood dynamics, on the other, are the two greatest challenges. Machine learning is providing cutting-edge, effective answers to all of these issues. Methods of reducing the number of dimensions are used to visualize dynamical processes in a more compact format. It is computationally efficient to use these methods on real -world data. Finding the fundamental equations fundamental principles of physics requires models that are themselves driven by evidence. By using deep learning techniques, dynamical systems are identified correctly, and that in turn enables accurate inference about physical systems. S everal cutting-edge and very successful new approaches to nonlinear dynamics have emerged from the area of machine learning. Autoencoders, recurrent neural networks, convolutional neural networks, and reinforcement learning are only some of the cutting-edge deep learning approaches used to simulate dynamical systems. © 2023 IEEE.

关键词： Autonomous vehicles

Unforgettable Lessons from Forgettable Images: Intra-Class Memorability Matters in computer Vision Tasks

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Jing, Jie Lin, Qing Han, Shuangpeng Schiatti, Lucia Kuo, Yen-Ling Zhang, Mengmi I2R CIFAR Agency for Science Technology and Research Singapore College of Computing and Data Science Nanyang Technological University Singapore Department of Computer Science Sichuan University China CSAIL CBMM Massachusetts Institute of Technology United States UVIP Istituto Italiano di Tecnologia Italy Department of Computer Science University of Virginia United States

We introduce intra-class memorability, where certain images within the same class are more memorable than others despite shared category characteristics. To investigate what features make one object instance more memorable than others, we design and conduct human behavior experiments, where participants are shown a series of images, and they must identify when the current image matches the image presented a few steps back in the sequence. To quantify memorability, we propose the Intra-Class Memorability score (ICMscore), a novel metric that incorporates the temporal intervals between repeated image presentations into its calculation. Furthermore, we curate the Intra-Class Memorability dataset (ICMD), comprising over 2,500 images across five object classes with their ICMscores derived from 1,000 participants’ responses. Subsequently, we demonstrate the usefulness of ICMD by training AI models on this dataset for various downstream tasks: memorability prediction, image recognition, continual learning, and memorability-controlled image editing. Surprisingly, high-ICMscore images impair AI performance in image recognition and continual learning tasks, while low-ICMscore images improve outcomes in these tasks. Additionally, we fine-tune a state-of-the-art image diffusion model on ICMD image pairs with and without masked semantic objects. The diffusion model can successfully manipulate image elements to enhance or reduce memorability. Our contributions open new pathways in (Figure Presented) Figure 1. An object instance can be more memorable than another within the same class. Intra-class memorability refers to the ability of instances within the same category to be remembered. First, we present example images with low and high memorability (first and second rows) for the exemplar flower class. A white pale daisy (Row 1) is less memorable than a striking violet osteospermum (Row 2). Images are identified as high or low memorability based on results from human behavior exp

关键词： Image recognition

An FL-based Energy-Efficient Opportunistic Intelligent Node Selection Scheme for Healthcare IoV Applications

学校读者我要写书评

暂无评论

An FL-based Energy-Efficient Opportunistic Intelligent Node ...

International Communication Systems and Networks and Workshops, COMSNETS

作者： Pallati Narsimhulu Rashmi Sahay Symbiosis Institute of Technology Hyderabad Campus Symbiosis International (Deemed University) Pune India Department of Computer Science and Engineering Faculty of Science and Technology (IcfaiTech) The ICFAI Foundation for Higher Education Hyderabad India Department of Data Science and Artificial Intelligence Faculty of Science and Technology (IcfaiTech) The ICFAI Foundation for Higher Education Hyderabad India

ISBN: (数字)9798331531195

ISBN: (纸本)9798331531201

The complexity of managing the energy and operational needs in large-scale Wireless Sensor Networks (WSNs) emphasizes the importance of energy-aware routing. This paper proposes an energy-efficient routing protocol that uses an intelligent node selection scheme to address energy-aware routing in the healthcare Internet of Vehicles (IoV). The protocol identifies the most suitable cluster leader based on dynamic criteria. By building on a ranking system that considers factors like location and remaining energy, the protocol selects the next intelligent node to establish the most energy-efficient route within the Internet of Things (IoT) context. The energy-aware routing scheme is tested across various clustering systems to evaluate its impact on route discovery based on system size and time. Experimental results show that the model reduces energy consumption during broadcast operations. Performance metrics such as Jitter and Packet Loss Ratio (PLR) further illustrate the protocol’s energy optimization. Simulation results across diverse network conditions demonstrate the superiority of adaptive ranking and intelligent forwarder node selection over competing strategies. Overall, the proposed protocol presents a promising solution for achieving energy-efficient routing in IoV, offering insights into optimized strategies and their benefits for IoT applications.

关键词： Wireless sensor networks Simulation Packet loss Medical services Jitter Routing Routing protocols Energy efficiency Vehicle dynamics Optimization

Joint Covert and Secure Communication for SWIPT Assisted CNOMA Systems

学校读者我要写书评

暂无评论

IEEE Internet of Things Journal 2025年第12期12卷 20406-20419页

作者： Huang, Gaojian Lei, Yuxin Li, Xingwang Khan, Wali Ullah Wang, Gongpu Nallanathan, Arumugam Henan Polytechnic University School of Physics and Electronic Information Engineering Jiaozuo454003 China City University of Macau Faculty of Data Science 999078 China Henan Polytechnic University School of Surveying and Land Information Engineering Jiaozuo454003 China Henan Polytechnic University School of Physics and Electronic Information Engineering Jiaozuo454099 China Southeast University National Mobile Communications Research Laboratory Nanjing210096 China Luxembourg City1855 Luxembourg Beijing Jiaotong University Beijing Key Laboratory of Transportation Data Analysis and Mining School of Computer and Information Technology Beijing100044 China Institute of Internet of Things Wuxi214111 China Queen Mary University of London School of Electronic Engineering and Computer Science LondonE1 4NS United Kingdom Kyung Hee University Department of Electronic Engineering Gyeonggi-do Yongin-si17104 Korea Republic of

With the rapid advancement of physical-layer security technology, the covert and secure communication has become crucial in safeguarding wireless communication systems. In this paper, we propose a joint covert and secure transmission scheme for simultaneous wireless information and power transfer (SWIPT) assisted cooperative non-orthogonal multiple access (CNOMA) systems. In the CNOMA system, a greedy relay transmits the confidential information to the far user (Carol), with the assistance of the near user (Bob). Meanwhile, as a SWIPT node, Bob is self-sustained by harvesting energy from relay. What's more, a warden (Alice) and non-colluding eavesdroppers (Eves) always attempt to detect and capture the confidential information, respectively. To counteract the attacks from Alice and Eves, a jamming-assisted scheme is employed. For the proposed system model, we derive closed-form expressions for the detection error probability (DEP) and the average minimum detection error probability (AMDEP) of Alice. Additionally, closed-form expressions for the outage probability (OP) of users and the intercept probability (IP) of Eves are obtained. Furthermore, to maximize the effective covert rate (ECR) of Carol, an optimization problem is formulated, subject to covertness and security constraints. Numerical results are provided to demonstrate the impact of the system parameters on covert and secure performance, with the results showing perfect agreement with the theoretical analysis. © 2014 IEEE.

关键词： Inductive power transmission