检索结果-内蒙古大学图书馆

Computational Intelligence 2023年第4期39卷 530-531页

作者： Er, Meng Joo Ju, Zhaojie Ferrein, Alexander Changjiang Scholar Distinguished Director of Institute of Artificial Intelligence and Marine Robotics College of Marine Electrical Engineering Dalian Maritime University Dalian China Reader in Machine Learning and Robotics School of Computing University of Portsmouth Portsmouth United Kingdom Department of Electrical Engineering and Information Technology Mobile Autonomous Systems & Cognitive Robotics Institute Robotics and Foundations of Computer Science FH Aachen - University of Applied Sciences Aachen Germany

来源：评论

学校读者我要写书评

暂无评论

PLM-based World Models for Text-based Games

PLM-based World Models for Text-based Games

引用

2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022

作者： Kim, Minsoo Jung, YeonJoon Lee, Dohyeon Hwang, Seung-Won Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of Department of Computer Science and Engineering Seoul National University Korea Republic of Department of Artificial Intelligence Yonsei University Korea Republic of

World models have improved the ability of reinforcement learning agents to operate in a sample efficient manner, by being trained to predict plausible changes in the underlying environment. As the core tasks of world models are future prediction and commonsense understanding, our claim is that pre-trained language models (PLMs) already provide a strong base upon which to build world models. Worldformer is a recently proposed world model for text-based game environments, based only partially on PLM and transformers. Our distinction is to fully leverage PLMs as actionable world models in text-based game environments, by reformulating generation as constrained decoding which decomposes actions into verb templates and objects. We show that our model improves future valid action prediction and graph change prediction. Additionally, we show that our model better reflects commonsense than standard PLM. © 2022 Association for Computational Linguistics.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Debiasing Classifiers by Amplifying Bias with Latent Diffusion and Large Language Models 25

Debiasing Classifiers by Amplifying Bias with Latent Diffusi...

引用

40th Annual ACM Symposium on Applied Computing, SAC 2025

作者： Ko, Donggeun Lee, Dongjun Park, Namjun Shim, Wonkyeong Kim, Jaekwang AiM Future Seoul Korea Republic of Maum AI Gyeonggi-do Korea Republic of Department of Computer Science and Engineering Convergence Program for Social Innovation Sungkyunkwan University Suwon Korea Republic of Department of Computer Science and Engineering Sungkyunkwan University Suwon Korea Republic of School of Convergence Convergence Program for Social Innovation Sungkyunkwan University Suwon Korea Republic of

ISBN: (纸本)9798400706295

Neural networks struggle with image classification when biases are learned and misleads correlations, affecting their generalization and performance. Previous methods require attribute labels (e.g. background, color) or utilizes Generative Adversarial Networks (GANs) to mitigate biases. We introduce DiffuBias, a novel pipeline for text-to-image generation that generates bias-conflict samples, without any training. By utilizing pretrained diffusion and image captioning models, DiffuBias generates, bias-conflict samples using the top-K losses from a biased classifier (fB) to debias the classifier. This method not only debiases effectively but also boosts classifier generalization capabilities. Our comprehensive experimental evaluations demonstrate that DiffuBias achieves state-of-the-art performance on benchmark datasets. Copyright © 2025 held by the owner/author(s).

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Spectral learning of shared dynamics between generalized-linear processes 24

Spectral learning of shared dynamics between generalized-lin...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Lucine L. Oganesian Omid G. Sani Maryam M. Shanechi Ming Hsieh Department of Electrical and Computer Engineering University of Southern California Los Angeles CA Ming Hsieh Department of Electrical and Computer Engineering Thomas Lord Department of Computer Science Alfred E. Mann Department of Biomedical Engineering Neuroscience Graduate Program University of Southern California Los Angeles CA

ISBN: (纸本)9798331314385

Generalized-linear dynamical models (GLDMs) remain a widely-used framework within neuroscience for modeling time-series data, such as neural spiking activity or categorical decision outcomes. Whereas the standard usage of GLDMs is to model a single data source, certain applications require jointly modeling two generalized-linear time-series sources while also dissociating their shared and private dynamics. Most existing GLDM variants and their associated learning algorithms do not support this capability. Here we address this challenge by developing a multi-step analytical subspace identification algorithm for learning a GLDM that explicitly models shared vs. private dynamics within two generalized-linear time-series. In simulations, we demonstrate our algorithm's ability to dissociate and model the dynamics within two time-series sources while being agnostic to their respective observation distributions. In neural data, we consider two specific applications of our algorithm for modeling discrete population spiking activity with respect to a secondary time-series. In both synthetic and real data, GLDMs learned with our algorithm more accurately decoded one time-series from the other using lower-dimensional latent states, as compared to models identified using existing GLDM learning algorithms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Innovations in Social Informatic: A Patent Landscape Review 7

Innovations in Social Informatic: A Patent Landscape Review

引用

7th Information Systems International Conference, ISICO 2023

作者： Afia, Nur Purnomo, Agung Maulana, Fairuz Iqbal Prasetyo, Yogi Tri Gumasing, Ma Janice J. Rosyidah, Elsa Persada, Satria Fadil Institute for Research and Community Services Universitas Nahdlatul Ulama Sidoarjo Sidoarjo61234 Indonesia English Language Education Universitas Nahdlatul Ulama Sidoarjo Sidoarjo61234 Indonesia Entrepreneurship Department BINUS Business School Undergraduate Program Bina Nusantara University Jakarta11480 Indonesia Computer Science Department Bina Nusantara University Jakarta11480 Indonesia School of Industrial Engineering and Engineering Management Mapúa University Manila1002 Philippines International Bachelor Program in Engineering Yuan Ze University Taoyuan32003 Taiwan

Patent intellectual property in social informatics with entrepreneurship support creates greater opportunities to improve human life. It has the potential to open enormous innovation opportunities in various fields, such as technology, health, services, and society. This research was conducted to review the patent landscape and the main trends of patents on social informatics in all countries. Researchers applied a patent landscape analysis using data on 487 patent documents over 16 years from the *** database. The study results show yearly patented development patterns related to social informatics. There is a royalty-free opportunity by using 15 patent discontinued. Patent related to social informatics was dominated by company ownership and inventors from the United States. The patent physics (G) section overwhelmed patents related to social informatics. There were 334 simple family patents on social informatics in nine jurisdictions;the United States was the highest. © 2023 The Authors. Published by Elsevier B.V.

关键词： Patents and inventions

来源：评论

学校读者我要写书评

暂无评论

Multi-Representation Genetic programming: A Case Study on Tree-based and Linear Representations

arXiv

引用

arXiv 2024年

作者： Huang, Zhixing Mei, Yi Zhang, Fangfang Zhang, Mengjie Banzhaf, Wolfgang Centre for Data Science and Artificial Intelligence School of Engineering and Computer Science Victoria University of Wellington Wellington6140 New Zealand Department of Computer Science and Engineering BEACON Center for the Study of Evolution in Action and Ecology Evolution and Behavior Program Michigan State University East LansingMI48864 United States

Existing genetic programming (GP) methods are typically designed based on a certain representation, such as tree-based or linear representations. These representations show various pros and cons in different domains. However, due to the complicated relationships among representation and fitness landscapes of GP, it is hard to intuitively determine which GP representation is the most suitable for solving a certain problem. Evolving programs (or models) with multiple representations simultaneously can alternatively search on different fitness landscapes since representations are highly related to the search space that essentially defines the fitness landscape. Fully using the latent synergies among different GP individual representations might be helpful for GP to search for better solutions. However, existing GP literature rarely investigates the simultaneous effective use of evolving multiple representations. To fill this gap, this paper proposes a multi-representation GP algorithm based on tree-based and linear representations, which are two commonly used GP representations. In addition, we develop a new cross-representation crossover operator to harness the interplay between tree-based and linear representations. Empirical results show that navigating the learned knowledge between basic tree-based and linear representations successfully improves the effectiveness of GP with solely tree-based or linear representation in solving symbolic regression and dynamic job shop scheduling problems. © 2024, CC BY.

关键词： Genetic programming

来源：评论

学校读者我要写书评

暂无评论

Forward and Reverse Converters for the Moduli-Set {22q+1, 2q + 2q−1 ± 1}

arXiv

引用

arXiv 2024年

作者： Jaberipur, Ghassem Nadimi, Bardia Kazemi, R. Lee, Jeong-A The Brain Pool Program for Chosun University Gwangju Korea Republic of The University of South Florida TampaFL United States Computer Engineering The Department of Computer Science and Engineering Shahid Beheshti University Tehran Iran The Department of Computer Engineering Chosun University Korea Republic of

Modulo-(2q + 2q−1 ± 1) adders have recently been implemented using the regular parallel prefix (RPP) architecture, matching the speed of the widely used modulo-(2q ± 1) RPP adders. Consequently, we introduce a new moduli set τ+ = {22q+1, 2q + 2q−1 ± 1}, with over (2q+2)× dynamic range and adder speeds comparable to the conventional τ = {2q, 2q ± 1} set. However, to fully leverage τ+ in residue number system applications, a complete set of circuitries is necessary. This work focuses on the design and implementation of the forward and reverse converters for τ+. These converters consist of four and seven levels of carry-save addition units, culminating in a final modulo-(2q + 2q−1 ± 1) and modulo-(22q+1 + 22q−2 − 1) adder, respectively. Through analytical evaluations and circuit simulations, we demonstrate that the overall performance of a sequence of operations-including residue generation, k additions, and reverse conversion-using τ+ surpasses that of τ when k exceeds a certain practical threshold. Copyright © 2024, The Authors. All rights reserved.

关键词： Adders

来源：评论

学校读者我要写书评

暂无评论

Design and Optimization of the Porous Metamaterial EM Absorbers in X-band

Design and Optimization of the Porous Metamaterial EM Absorb...

引用

IEEE International Symposium on Electromagnetic Compatibility (EMC)

作者： Kanat Anurakparadorn Alan Taub Eric Michielssen Macromolecular Science and Engineering Program Department of Material Science and Engineering University of Michigan Ann Arbor Michigan USA Department of Mechanical Engineering University of Michigan Ann Arbor Michigan USA Department of Electrical Engineering and Computer Science University of Michigan Ann Arbor Michigan USA

A strategy that combines experiment and simulation to design and optimize electromagnetic (EM) metamaterial absorbers containing a periodic porous structure is described. The approach provides the ability to produce absorbers that meet multiple user-specified objectives. Using the measured intrinsic properties of the baseline materials as an input to EM-field based computational modelling and optimization, absorption by the studied metamaterials measured by their reflection loss (RL) increases significantly. The resulting metamaterials have the potential for lower cost and lighter weight while providing greater protection than traditional metal gaskets and foams.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning Deep Tree-based Retriever for Efficient Recommendation: Theory and Method

arXiv

引用

arXiv 2024年

作者： Liu, Ze Zhang, Jin Feng, Chao Lian, Defu Wang, Jie Chen, Enhong The School of Computer Science and Technology University of Science and Technology of China Anhui Hefei230027 China The School of Artificial Intelligence and Data Science University of Science and Technology of China Anhui Hefei230027 China The State Key Laboratory of Cognitive Intelligence School of Computer Science and Technology School of Artificial Intelligence and Data Science University of Science and Technology of China Anhui Hefei230027 China The Department of Electronic Engineering and Information Science University of Science and Technology of China Anhui Hefei230027 China

With the advancement of deep learning, deep recommendation models have achieved remarkable improvements in recommendation accuracy. However, due to the large number of candidate items in practice and the high cost of preference computation, these methods still suffer from low recommendation efficiency. The recently proposed tree-based deep recommendation models alleviate the problem by directly learning tree structure and representations under the guidance of recommendation objectives. To guarantee the effectiveness of beam search for recommendation accuracy, these models strive to ensure that the tree adheres to the max-heap assumption, where a parent node's preference should be the maximum among its children's preferences. However, they employ a one-versus-all strategy, framing the training task as a series of independent binary classification objectives for each node, which limits their ability to fully satisfy the max-heap assumption. To this end, we propose a Deep Tree-based Retriever (DTR for short) for efficient recommendation. DTR frames the training task as a softmax-based multi-class classification over tree nodes at the same level, enabling explicit horizontal competition and more discriminative top-k selection among them, which mimics the beam search behavior during training. To mitigate the suboptimality induced by the labeling of non-leaf nodes, we propose a rectification method for the loss function, which further aligns with the max-heap assumption in expectation. As the number of tree nodes grows exponentially with the levels, we employ sampled softmax to approximate optimization and thereby enhance efficiency. Furthermore, we propose a tree-based sampling method to reduce the bias inherent in sampled softmax. Theoretical results reveal DTR's generalization capability, and both the rectification method and tree-based sampling contribute to improved generalization. The experiments are conducted on four real-world datasets, validating the effectivenes

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Learning to Escape: Multi-mode Policy Learning for the Traveling Salesmen Problem

Learning to Escape: Multi-mode Policy Learning for the Trave...

引用

IEEE Workshop on Evolving and Adaptive Intelligent Systems (EAIS)

作者： Myoung Hoon Ha Seunggeun Chi Sang Wan Lee Center for Neuroscience-inspired AI KAIST Daejeon Korea School of Electrical & Computer Engineering Purdue University West Lafayette USA Department of Brain Cognitive Science KAIST Daejeon Korea

ISBN: (数字)9798350366235

ISBN: (纸本)9798350366242

The traveling salesmen problem (TSP)-one of the most fundamental NP-hard problems in combinatorial optimization-has received considerable attention owing to its direct applicability to real-world routing. Recent studies on TSP have adopted a deep policy network to learn a stochastic acceptance rule. Despite its success in some cases, the structural and functional complexity of the deep policy networks makes it hard to explore the problem space while performing a local search at the same time. We found in our empirical analyses that searching processes are often stuck in the local region, leading to severe performance degradation. To tackle this issue, we propose a novel method for multi-mode policy learning. In the proposed method, a conventional exploration-exploitation scheme is reformulated as the problem of learning to escape from a local search area to induce exploration. We present a multi-mode Markov decision process, followed by policy and value design for local search and escaping modes. Experimental results show that the performance of the proposed method is superior to that of various baseline models, suggesting that the learned escaping policy allows the model to initiate a new local search in promising regions efficiently.

关键词： Degradation NP-hard problem Markov decision processes Stochastic processes Search problems Routing Space exploration

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：