2021 International Joint Conference on Neural Networks (IJCNN)最新文献_第2页

Layerwise Approximate Inference for Bayesian Uncertainty Estimates on Deep Neural Networks 深度神经网络贝叶斯不确定性估计的分层近似推理

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534229

Ni Zhang, Xiaoyi Chen, Li Quan

{"title":"Layerwise Approximate Inference for Bayesian Uncertainty Estimates on Deep Neural Networks","authors":"Ni Zhang, Xiaoyi Chen, Li Quan","doi":"10.1109/IJCNN52387.2021.9534229","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9534229","url":null,"abstract":"A proper representation of predictive uncertainty is vital for deep neural networks (DNNs) to be applied in safety-critical domains such as medical diagnosis and self-driving. State-of-the-art (SOTA) variational inference approximation techniques provide a theoretical framework for modeling uncertainty, however, they have not been proven to work on large and deep networks with practical computation. In this study, we develop a layerwise approximation with a local reparameterization technique to efficiently perform sophisticated variational Bayesian inference on very deep SOTA convolutional neural networks (CNNs) (VGG16, ResNet variants, DenseNet). Theoretical analysis is presented to justify that the layerwise approach remains a Bayesian neural network. We further derive a SOTA $alpha$-divergence objective function to work with the layerwise approximate inference, addressing the concern of underestimating uncertainties by the Kullback-Leibler divergence. Empirical evaluation using MNIST, CIFAR-10, and CIFAR-100 datasets consistently shows that with our proposal, deep CNN models can have a better quality of predictive uncertainty than Monte Carlo-dropout in detecting in-domain misclassification and excel in out-of-distribution detection.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121240807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Exploring the Effect of Dynamic Drive Balancing in Open-ended Learning Robots 探讨动态驱动平衡在开放式学习机器人中的作用

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534137

A. Romero, F. Bellas, R. Duro

{"title":"Exploring the Effect of Dynamic Drive Balancing in Open-ended Learning Robots","authors":"A. Romero, F. Bellas, R. Duro","doi":"10.1109/IJCNN52387.2021.9534137","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9534137","url":null,"abstract":"This paper seeks to explore the effect and possibilities of autonomously balancing drives in a motivational architecture aimed at open-ended learning robots. These types of robots are very useful in unconstrained human robot interaction settings or when uncontrolled dynamic scenarios that are unknown at design time must be addressed. Designing a robot under these conditions implies that it must be endowed with some primary operational purpose and some additional self-preservation objectives whose fulfillment depend on the characteristics of the particular domain it is facing each moment in time. Domains that are not known beforehand and for which no a priori goal or skill structure can be designed in. Thus, an approach to the design and engineering of motivational structures to endow robots with specific purposes is proposed and tested here. We concentrate on the drive structure of a motivational system and the effects of its autonomous adaptation to changing circumstances. To provide for this adaptation, a simple evolutionary strategy is defined for the autonomous regulation of multiple drives seeking to optimize long-term operation. The proposal is tested on a Baxter robot performing an industrial task and the results confirm the potential of autonomous dynamic drive balancing as a tool in open-ended settings.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127465895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Structure and Randomness in Planning and Reinforcement Learning 计划和强化学习中的结构和随机性

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9533317

K. Czechowski, Piotr Januszewski, Piotr Kozakowski, Łukasz Kuciński, Piotr Milos

引用次数: 3

Bag of Tricks for Chinese Named Entity Recognition 中文命名实体识别技巧包

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9533296

Yao Xiao, Jingbo Peng, Luoyi Fu, Haisong Zhang

引用次数: 0

Multi-Agent Combat in Non-Stationary Environments 非静止环境下的多智能体战斗

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534036

Shengang Li, Haoang Chi, Tao Xie

{"title":"Multi-Agent Combat in Non-Stationary Environments","authors":"Shengang Li, Haoang Chi, Tao Xie","doi":"10.1109/IJCNN52387.2021.9534036","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9534036","url":null,"abstract":"Multi-agent combat is a combat scenario in multiagent reinforcement learning (MARL). In this combat, agents use reinforcement learning methods to learn optimal policies. Actually, policy may be changed, which leads to a non-stationary environment. In this case, it is difficult to predict opponents' policies. Many reinforcement learning methods try to solve nonstationary problems. Most of the previous works put all agents into a frame and model their policies to deal with non-stationarity of environments. But, in a combat environment, opponents can not be in the same frame as our agents. We group opponents and our agents into two frames, only considering opponents as a part of the environment. In this paper, we focus on the problem of modelling opponents' policies in non-stationary environments. To solve this problem, we propose an algorithm called Additional Opponent Characteristics Multi-agent Deep Deterministic Policy Gradient (AOC-MADDPG) with the following contributions: (1) We propose a new actor-critic framework to deal with nonstationarity of environments in MARL, so that agents can adapt to more complex environments. (2) A model for opponents' policies is built by introducing observations and actions of the opponents into the critic network as additional characteristics. We evaluate our AOC-MADDPG algorithm in two multi-agent combat environments. As a result, our approach significantly outperforms the baseline. Agents trained by our method can get higher rewards in non-stationary environments.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"456 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125793357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

A Self-Supervised Learning Framework for Sequential Recommendation

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534405

Renqi Jia, Xu Bai, Xiaofei Zhou, Shirui Pan

{"title":"A Self-Supervised Learning Framework for Sequential Recommendation","authors":"Renqi Jia, Xu Bai, Xiaofei Zhou, Shirui Pan","doi":"10.1109/IJCNN52387.2021.9534405","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9534405","url":null,"abstract":"Sequential recommendation that aims to predict user preference with historical user interactions becomes one of the most popular tasks in the recommendation area. The existing methods concentrated on user's sequential features among exposed items have achieved good performance. However, they only rely on single item prediction optimization to learn data representation, which ignores the association between context data and sequence data. In this paper, we propose a novel self-supervised learning based sequential recommendation network (SSLRN), which contrastively learns data correlation to promote data representation of users and items. We design two auxiliary contrastive learning tasks to regularize user and item representation based on mutual information maximization (MIM). In particular, the item contrastive learning captures sequential contrast feature with sequence-item MIM, and the user contrastive learning regularizes user latent representation with user-item MIM. We evaluate our model on five real-world datasets and the experimental results show that the proposed framework significantly and consistently outperforms state-of-the-art sequential recommendation techniques.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125842114","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Prior Probability Estimation in Dynamically Imbalanced Data Streams 动态不平衡数据流中的先验概率估计

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9533795

Joanna Komorniczak, Paweł Zyblewski, Paweł Ksieniewicz

引用次数: 6

Mastering the Game of Amazons Fast by Decoupling Network Learning 通过解耦网络学习快速掌握亚马逊游戏

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534274

G. Q. Zhang, Xiaoyang Chen, Ruidong Chang, Yuhang Zhang, Cong Wang, Luyi Bai, Junwei Wang, Changming Xu

引用次数: 1

Graph-based Visual Manipulation Relationship Reasoning in Object-Stacking Scenes 对象堆叠场景中基于图的视觉操作关系推理

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9534389

Guoyu Zuo, Jiayuan Tong, Hongxing Liu, Wenbai Chen, Jianfeng Li

{"title":"Graph-based Visual Manipulation Relationship Reasoning in Object-Stacking Scenes","authors":"Guoyu Zuo, Jiayuan Tong, Hongxing Liu, Wenbai Chen, Jianfeng Li","doi":"10.1109/IJCNN52387.2021.9534389","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9534389","url":null,"abstract":"In object-stacking scenes, robotic manipulation is one of the most important research topics in robotics. It is particularly significant to reason object relationships and obtain intelligent manipulation order for more advanced interaction between the robot and the environment. However, many existing methods focus on individual object features and ignore contextual information, leading to great challenges in efficiently reasoning manipulation relationship. In this paper, we introduce a novel graph-based visual manipulation relationship reasoning architecture that directly outputs object relationships and manipulation order. Our model first extracts features and detects objects from RGB images, and then adopts Graph Convolutional Network (GCN) to collect contextual information between objects. Moreover, a relationship filtering network is built to reduce object pairs before reasoning and improve the efficiency of relation reasoning. The experiments on the Visual Manipulation Relationship Dataset (VMRD) show that our model significantly outperforms previous methods on reasoning object relationships in obiect-stackina scenes.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126076786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

A Strategic Weight Refinement Maneuver for Convolutional Neural Networks 卷积神经网络的一种策略权重细化策略

2021 International Joint Conference on Neural Networks (IJCNN) Pub Date : 2021-07-18 DOI: 10.1109/IJCNN52387.2021.9533359

Patrick Sharma, Adarsh Karan Sharma, Dinesh Kumar, Anuraganand Sharma

{"title":"A Strategic Weight Refinement Maneuver for Convolutional Neural Networks","authors":"Patrick Sharma, Adarsh Karan Sharma, Dinesh Kumar, Anuraganand Sharma","doi":"10.1109/IJCNN52387.2021.9533359","DOIUrl":"https://doi.org/10.1109/IJCNN52387.2021.9533359","url":null,"abstract":"Stochastic Gradient Descent algorithms (SGD) remain a popular optimizer for deep learning networks and have been increasingly used in applications involving large datasets producing promising results. SGD approximates the gradient on a small subset of training examples, randomly selected in every iteration during network training. This randomness leads to the selection of an inconsistent order of training examples resulting in ambiguous values to solve the cost function. This paper applies Guided Stochastic Gradient Descent (GSGD) - a variant of SGD in deep learning neural networks. GSGD minimizes the training loss and maximizes the classification accuracy by overcoming the inconsistent order of data examples in SGDs. It temporarily bypasses the inconsistent data instances during gradient computation and weight update, leading to better convergence at the rate of $O(frac{1}{rho T-})$. Previously, GSGD has only been used in the shallow learning networks like the logistic regression. We try to incorporate GSGD in deep learning neural networks like the Convolutional Neural Networks (CNNs) and evaluate the classification accuracy in comparison with the same networks trained with SGDs. We test our approach on benchmark image datasets. Our baseline results show GSGD leads to a better convergence rate and improves classification accuracy by up to 3% of standard CNNs.","PeriodicalId":396583,"journal":{"name":"2021 International Joint Conference on Neural Networks (IJCNN)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126109700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1