39,039 Hits in 8.6 sec

Modelling collective motion based on the principle of agency: General framework and the case of marching locusts

Katja Ried, Thomas Müller, Hans J. Briegel, Iman Borazjani
2019 PLoS ONE  
We propose a new approach to studying collective behaviour, based on the concept of learning agents: individuals endowed with explicitly modelled sensory capabilities, an internal mechanism for deciding  ...  We illustrate these points with the example of marching locusts, showing how learning agents can account for the phenomenon of density-dependent alignment.  ...  In the present work, we propose a more comprehensive framework for modelling collective motion, based on the notion of learning agents (or simply 'agents', for short): entities that interact with their  ... 
doi:10.1371/journal.pone.0212044 pmid:30785947 pmcid:PMC6382133 fatcat:uhqstqcvlzcppk5xay6hhv2nwu

Guest Editorial: Special Issue on AI-Enabled Internet of Dependable and Controllable Things

Wei Yu, Wei Zhao, Anke Schmeink, Houbing Song, Guido Dartmann
2021 IEEE Internet of Things Journal  
Furthermore, to optimize the alignment of network entities, a group structure aggregation optimization module is developed.  ...  The article titled "Dynamic Bayesian collective awareness models for network of ego-things" proposes to learn collective awareness models from low-dimensional data sequences of a network with intelligent  ... 
doi:10.1109/jiot.2021.3053713 fatcat:wnsgkuohhvg4fitk6ixreddsly

Reinforced Iterative Knowledge Distillation for Cross-Lingual Named Entity Recognition [article]

Shining Liang, Ming Gong, Jian Pei, Linjun Shou, Wanli Zuo, Xianglin Zuo, Daxin Jiang
2021 arXiv   pre-print
To effectively extract weak supervision signals from the unlabeled data, we develop a novel approach based on the ideas of semi-supervised learning and reinforcement learning.  ...  Named entity recognition (NER) is a fundamental component in many applications, such as Web Search and Voice Assistants.  ...  [6, 14] leverage extra knowledge base or word alignment tools to annotate training data. [16, 25] generate target-language training data with machine translation.  ... 
arXiv:2106.00241v1 fatcat:mvayp27cy5hc5gwefgan3ynr3e

Adaptive Behavior Generation for Autonomous Driving using Deep Reinforcement Learning with Compact Semantic States [article]

Peter Wolf, Karl Kurzer, Tobias Wingert, Florian Kuhnt, J. Marius Zöllner
2018 arXiv   pre-print
Therefore it is hard to model solely based on expert knowledge. In this work we use Deep Reinforcement Learning to learn maneuver decisions based on a compact semantic state representation.  ...  With little expert knowledge and a set of mid-level actions, it can be seen that the agent is capable to adhere to traffic rules and learns to drive safely in a variety of situations.  ...  APPROACH We employ a deep reinforcement learning approach to generate adaptive behavior for autonomous driving.  ... 
arXiv:1809.03214v1 fatcat:u55npj4ot5ffni7edmcfs373uq

Graph Adaptation Network with Domain-Specific Word Alignment for Cross-Domain Relation Extraction

Zhe Wang, Bo Yan, Chunhua Wu, Bin Wu, Xiujuan Wang, Kangfeng Zheng
2020 Sensors  
Most existing works adapted relation extraction models from the source domain to target domain through aligning sequential features, but failed to transfer non-local and non-sequential features such as  ...  To address this issue, in this paper, we propose a novel tripartite graph architecture to adapt non-local features when there is no labeled data in the target domain.  ...  In the last decade, deep learning (DL) has made breakthroughs in natural language processing (NLP), image processing and reinforcement learning, so that making breakthroughs in the field of artificial  ... 
doi:10.3390/s20247180 pmid:33333844 fatcat:bjjnzbqdabfspmpfl7nkabtn7a

Multi-source knowledge fusion: a survey

Xiaojuan Zhao, Yan Jia, Aiping Li, Rong Jiang, Yichen Song
2020 World wide web (Bussum)  
Due to the uncertainty of knowledge acquisition, the reliability and confidence of KG based on entity recognition and relationship extraction technology need to be evaluated.  ...  On this basis, the challenges and future research directions of multisource knowledge fusion in a large-scale knowledge base environment are discussed.  ...  Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long  ... 
doi:10.1007/s11280-020-00811-0 fatcat:ef5j2sna6fai7k2455yihrrfuq

Federated Transfer Learning: concept and applications [article]

Sudipan Saha, Tahir Ahmad
2021 arXiv   pre-print
Among different variants of the federated learning, noteworthy is federated transfer learning (FTL) that allows knowledge to be transferred across domains that do not have many overlapping features and  ...  However, in most industries data exists in form of isolated islands, with limited scope of sharing between different organizations. This is an hindrance to the further development of AI.  ...  For feature alignment between source and target, correlation alignment (CORAL) loss is used, which adapts the second order feature statistics [32] .  ... 
arXiv:2010.15561v3 fatcat:3udixrhta5btlb7w7r4fomwpzu

Towards Robot Task Planning From Probabilistic Models of Human Skills [article]

Chris Paxton, Marin Kobilarov, Gregory D. Hager
2016 arXiv   pre-print
We represent robot skills in terms of a probability distribution over features learned from multiple expert demonstrations.  ...  We describe an algorithm for motion planning based on expert demonstrations of a skill.  ...  Kormushev et al. used a modified version of DMPs together with reinforcement learning to adapt to new environments with a known goal [14] .  ... 
arXiv:1602.04754v1 fatcat:gjjv4z6ybjg2dcknt3bqcyzkxq

Pairing Conceptual Modeling with Machine Learning [article]

Wolfgang Maass, Veda C. Storey
2021 arXiv   pre-print
With the increasing emphasis on digitizing and processing large amounts of data for business and other applications, it would be helpful to consider how these areas of research can complement each other  ...  Both conceptual modeling and machine learning have long been recognized as important areas of research.  ...  Acknowledgements This paper was based on a keynote presentation given by the first author at the International Conference on Conceptual Modeling.  ... 
arXiv:2106.14251v1 fatcat:n4kujuzttja67jqjs3vz3bdiba

K-Adapter: Infusing Knowledge into Pre-Trained Models with Adapters [article]

Ruize Wang, Duyu Tang, Nan Duan, Zhongyu Wei, Xuanjing Huang, Jianshu ji, Guihong Cao, Daxin Jiang, Ming Zhou
2020 arXiv   pre-print
Results on three knowledge-driven tasks, including relation classification, entity typing, and question answering, demonstrate that each adapter improves the performance and the combination of both adapters  ...  As a case study, we inject two kinds of knowledge in this work, including (1) factual knowledge obtained from automatically aligned text-triplets on Wikipedia and Wikidata and (2) linguistic knowledge  ...  They align entities from Wikipedia sentences to fact triples in WikiData, and discard sentences with less than three entities.  ... 
arXiv:2002.01808v5 fatcat:cqyzyiuljrgitkz27fs4jefcg4

Putting Humans in the Natural Language Processing Loop: A Survey [article]

Zijie J. Wang, Dongjin Choi, Shenyu Xu, Diyi Yang
2021 arXiv   pre-print
HITL NLP research is nascent but multifarious -- solving various NLP problems, collecting diverse feedback from different people, and applying different methods to learn from collected feedback.  ...  How can we design Natural Language Processing (NLP) systems that learn from human feedback?  ...  To tackle this challenge, online reinforcement learning can be used to improve the model with human feedback.  ... 
arXiv:2103.04044v1 fatcat:bnwj25lwofcwrnjtvlta64niq4

Inductive Unsupervised Domain Adaptation for Few-Shot Classification via Clustering [article]

Xin Cong, Bowen Yu, Tingwen Liu, Shiyao Cui, Hengzhu Tang, Bin Wang
2020 arXiv   pre-print
Alignment, which are combined with a Cosine Annealing Strategy.  ...  In order to derive high-quality pseudo labels, we propose a Clustering Promotion Mechanism, to learn better features for the target domain via Similarity Entropy Minimization and Adversarial Distribution  ...  An example comes from FewRel 2.0, a few-shot dataset for relation classification with domain adaptation. Different colors indicate different entities, red for head entity, and blue for tail entity.  ... 
arXiv:2006.12816v1 fatcat:duwegr3z45ennfb3j6zk5zwdji

Adaptive Generation in Dialogue Systems Using Dynamic User Modeling

Srinivasan Janarthanam, Oliver Lemon
2014 Computational Linguistics  
We approach this problem using a three-step process: collecting data using a Wizard-of-Oz method, building simulated users, and learning to model and adapt to users using Reinforcement Learning techniques  ...  As an example, we show how a system can learn to choose referring expressions to refer to domain entities for users with different levels of domain expertise, and whose domain knowledge is initially unknown  ...  We present a three-step process to learning user-adaptive behavior in dialogue systems: data collection, building user simulations, and learning adaptive behavior using reinforcement learning.  ... 
doi:10.1162/coli_a_00203 fatcat:3v3wetxiabd3zhtmta3zsdflvq

Towards Cooperative Self-adapting Activity Recognition

Andreas Jahn, Sven Tomforde, Michel Morold, Klaus David, Bernhard Sick
2018 Proceedings of the 8th International Joint Conference on Pervasive and Embedded Computing and Communication Systems  
Although being a well-established research field, several basic issues are still insufficiently solved, including extensibility of an AR system at runtime, adaption of classification models to a very specific  ...  To overcome these limitations, the cooperation of AR systems including sporadic interaction with humans and consideration of other information sources is proposed in this article as a basic new way to  ...  with reinforcement learning techniques.  ... 
doi:10.5220/0006856102150222 dblp:conf/peccs/JahnTMDS18 fatcat:jqd5nvb5zjhdhbn5vyjeh4v7tq

GroupCap: Group-Based Image Captioning with Structured Relevance and Diversity Constraints

Fuhai Chen, Rongrong Ji, Xiaoshuai Sun, Yongjian Wu, Jinsong Su
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
Finally, such correlations are modeled as constraints and sent into the LSTM-based captioning generator.  ...  We adopt an end-to-end formulation to train the visual tree parser, the structured relevance and diversity constraints, as well as the LSTM based captioning model jointly.  ...  To construct this dataset, we firstly filter and collect the top-784 entities and the top-246 relations with high frequencies in the textual parsing tree.  ... 
doi:10.1109/cvpr.2018.00146 dblp:conf/cvpr/ChenJSWS18 fatcat:soi74rhcgrdrlk3wmzrunyuxl4
« Previous Showing results 1 — 15 out of 39,039 results