Applied Intelligence最新文献

Learning state-action correspondence across reinforcement learning control tasks via partially paired trajectories

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-06190-7

Javier García, Iñaki Rañó, J. Miguel Burés, Xosé R. Fdez-Vidal, Roberto Iglesias

{"title":"Learning state-action correspondence across reinforcement learning control tasks via partially paired trajectories","authors":"Javier García, Iñaki Rañó, J. Miguel Burés, Xosé R. Fdez-Vidal, Roberto Iglesias","doi":"10.1007/s10489-024-06190-7","DOIUrl":"10.1007/s10489-024-06190-7","url":null,"abstract":"<div><p>In many reinforcement learning (RL) tasks, the state-action space may be subject to changes over time (e.g., increased number of observable features, changes of representation of actions). Given these changes, the previously learnt policy will likely fail due to the mismatch of input and output features, and another policy must be trained from scratch, which is inefficient in terms of <i>sample complexity</i>. Recent works in transfer learning have succeeded in making RL algorithms more efficient by incorporating knowledge from previous tasks, thus partially alleviating this problem. However, such methods typically must provide an explicit state-action correspondence of one task into the other. An autonomous agent may not have access to such high-level information, but should be able to analyze its experience to identify similarities between tasks. In this paper, we propose a novel method for automatically learning a correspondence of states and actions from one task to another through an agent’s experience. In contrast to previous approaches, our method is based on two key insights: i) only the first state of the trajectories of the two tasks is <i>paired</i>, while the rest are <i>unpaired</i> and randomly collected, and ii) the transition model of the source task is used to predict the dynamics of the target task, thus aligning the <i>unpaired</i> states and actions. Additionally, this paper intentionally decouples the learning of the state-action corresponce from the transfer technique used, making it easy to combine with any transfer method. Our experiments demonstrate that our approach significantly accelerates transfer learning across a diverse set of problems, varying in state/action representation, physics parameters, and morphology, when compared to state-of-the-art algorithms that rely on cycle-consistency.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 3","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A domain-aware model with multi-perspective contrastive learning for natural language understanding

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-06154-x

Di Wang, Qingjian Ni

{"title":"A domain-aware model with multi-perspective contrastive learning for natural language understanding","authors":"Di Wang, Qingjian Ni","doi":"10.1007/s10489-024-06154-x","DOIUrl":"10.1007/s10489-024-06154-x","url":null,"abstract":"<div><p>Intent detection and slot filling are core tasks in natural language understanding (NLU) for task-oriented dialogue systems. However, current models face challenges with numerous intent categories, slot types, and domain classifications, alongside a shortage of well-annotated datasets, particularly in Chinese. Therefore, we propose a domain-aware model with multi-perspective, multi-positive contrastive learning. First, we adopt a self-supervised contrastive learning with multiple perspectives and multiple positive instances, which is capable of spacing the vectors of positive and negative instances from the domain, intent, and slot perspectives, and fusing more positive instance information to increase the classification effectiveness of the model. Our proposed domain-aware model defines domain-level units at the decoding layer, allowing the model to predict intent and slot information based on domain features, which greatly reduces the search space for intent and slot. In addition, we design a dual-stage attention mechanism for capturing implicitly shared information between intents and slots. We propose a data augmentation method that adds noise to the embedding layer, applies fine-grained augmentation techniques, and filters biased samples based on a similarity threshold. Our model is applied to real task-oriented dialogue systems and compared with other NLU models. Experimental results demonstrate that our proposed model outperforms other models in terms of NLU performance.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 3","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Stabilizing and improving federated learning with highly non-iid data and client dropout

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-05956-3

Jian Xu, Meilin Yang, Wenbo Ding, Shao-Lun Huang

{"title":"Stabilizing and improving federated learning with highly non-iid data and client dropout","authors":"Jian Xu, Meilin Yang, Wenbo Ding, Shao-Lun Huang","doi":"10.1007/s10489-024-05956-3","DOIUrl":"10.1007/s10489-024-05956-3","url":null,"abstract":"<div><p>The label distribution skew has been shown to be a significant obstacle that limits the model performance in federated learning (FL). This challenge could be more serious when the participating clients are in unstable network circumstances and drop out frequently. Previous works have demonstrated that the classifier head is particularly sensitive to the label skew. Therefore, maintaining a balanced classifier head is of significant importance for building a good and unbiased global model. To this end, we propose a simple yet effective framework by introducing a calibrated softmax function with smoothed prior for computing the cross-entropy loss, and a prototype-based feature augmentation scheme to re-balance the local training, which provide a new perspective on tackling the label distribution skew in FL and are lightweight for edge devices and can facilitate the global model aggregation. With extensive experiments on two benchmark classification tasks of Fashion-MNIST and CIFAR-10, our numerical results demonstrate that our proposed method can consistently outperform the baselines, 2 8% of accuracy over FedAvg in the presence of severe label skew and client dropout.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 3","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875251","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Efficient knowledge distillation using a shift window target-aware transformer

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-06207-1

Jing Feng, Wen Eng Ong

引用次数: 0

An attribute reduction algorithm using relative decision mutual information in fuzzy neighborhood decision system

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-06171-w

Jiucheng Xu, Shan Zhang, Miaoxian Ma, Wulin Niu, Jianghao Duan

{"title":"An attribute reduction algorithm using relative decision mutual information in fuzzy neighborhood decision system","authors":"Jiucheng Xu, Shan Zhang, Miaoxian Ma, Wulin Niu, Jianghao Duan","doi":"10.1007/s10489-024-06171-w","DOIUrl":"10.1007/s10489-024-06171-w","url":null,"abstract":"<div><p>The fuzzy neighborhood rough set integrates the strengths of fuzzy rough set and neighborhood rough set, serving as a pivotal extension of the rough set theory in attribute reduction. However, this model’s widespread application is hindered by its sensitivity to data distribution and limited efficacy in assessing classification uncertainty for datasets with substantial density variations. To mitigate these challenges, this paper introduces an attribute reduction algorithm based on fuzzy neighborhood relative decision mutual information. Firstly, the classification uncertainty of samples is initially defined in terms of relative distance. Simultaneously, the similarity relationship of fuzzy neighborhoods is reformulated, thereby reducing the risk of sample misclassification through integration with variable-precision fuzzy neighborhood rough approximation. Secondly, the notion of representative sample is introduced, leading to a redefinition of fuzzy membership. Thirdly, fuzzy neighborhood relative mutual information from the information view is constructed and combined with fuzzy neighborhood relative dependency from the algebraic view to propose fuzzy neighborhood relative decision mutual information. Finally, an attribute reduction algorithm is devised based on fuzzy neighborhood relative decision mutual information. This algorithm evaluates the significance of attributes by integrating both informational and algebraic perspectives. Comparative tests on 12 public datasets are conducted to assess existing attribute approximation algorithms. The experimental results show that the proposed algorithm achieved an average classification accuracy of 91.28<span>(%)</span> with the KNN classifier and 89.86<span>(%)</span> with the CART classifier. In both classifiers, the algorithm produced an average reduced subset size of 8.54. While significantly reducing feature redundancy, the algorithm consistently maintains a high level of classification accuracy.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 3","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875248","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Short-term traffic flow prediction based on spatial–temporal attention time gated convolutional network with particle swarm optimization

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-24 DOI: 10.1007/s10489-024-06117-2

Zhongxing Li, Zenan Li, Chaofeng Pan, Jian Wang

{"title":"Short-term traffic flow prediction based on spatial–temporal attention time gated convolutional network with particle swarm optimization","authors":"Zhongxing Li, Zenan Li, Chaofeng Pan, Jian Wang","doi":"10.1007/s10489-024-06117-2","DOIUrl":"10.1007/s10489-024-06117-2","url":null,"abstract":"<div><p>Recently, the surge in vehicle ownership has led to a corresponding increase in the complexity of traffic data. Consequently, accurate traffic flow prediction has become crucial for effective traffic management. While the advancements in intelligent transportation system (ITS) and internet of things (IoT) technology have facilitated traffic flow prediction, many existing methods overlook the influence of the training process on model accuracy. Traditional approaches often fail to account for this critical aspect. Hence, a new approach to traffic flow prediction is introduced in this paper: a spatial–temporal attention time-gated convolutional network based on particle swarm optimization (PSO-STATG). This method uses the particle swarm algorithm to dynamically optimize the learning rate and epoch parameters throughout the training process. Firstly, spatial–temporal correlations are extracted through spatial map convolution and time-gated convolution, facilitated by an attention mechanism. Subsequently, the learning rate and epoch parameters are dynamically adjusted during the training phase via the particle swarm optimization algorithm. Finally, experiments are conducted with real-world datasets, and the results are compared with those from several existing methods. The experimental results indicate that the accuracy and stability of our proposed model in predicting traffic flow are superior.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

CSRP: Modeling class spatial relation with prototype network for novel class discovery

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-23 DOI: 10.1007/s10489-024-05946-5

Wei Jin, Nannan Li, Jiuqing Dong, Huiwen Guo, Wenmin Wang, Chuanchuan You

{"title":"CSRP: Modeling class spatial relation with prototype network for novel class discovery","authors":"Wei Jin, Nannan Li, Jiuqing Dong, Huiwen Guo, Wenmin Wang, Chuanchuan You","doi":"10.1007/s10489-024-05946-5","DOIUrl":"10.1007/s10489-024-05946-5","url":null,"abstract":"<div><p>Novel Class Discovery(NCD) is a learning paradigm within the open-world task, in which machine learning models leverage prior knowledge to guide unknown samples into semantic clusters in an unsupervised environment. Recent research notes that maintaining class relations can assist classifiers in better recognizing unknown classes. Inspired by this study, we propose Class-Spatial-Relation modeling with a Prototype network (CSRP). A prototype network is a machine learning model used to classify tasks. It performs by learning prototypes for each class and makes classification decisions based on the similarity between a given sample and these prototypes. It conducts complex class boundaries better than linear classification models, providing higher flexibility and accuracy for classification tasks. Specifically, the proposed prototype network enables spatial modeling based on the distance between samples and each prototype, which can better obtain class relation information to improve the model’s interpretability and robustness. In addition, we simultaneously perform knowledge distillation on known and unknown classes to balance the model’s classification performance for each class. To evaluate the effectiveness and generality of our method, we perform extensive experiments on the CIFAR-100 dataset and fine-grained datasets: Stanford Cars, CUB-200-2011, and FGVC-Aircraft, respectively. Our method results are comparable to existing state-of-the-art performance in the standard dataset CIFAF100, while outstanding performance on three fine-grained datasets surpassed the baseline by 3%-9%. In addition, our method creates more compact clusters in the latent space than in linear classification. The success demonstrates the effectiveness of our approach.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142870458","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SPECN:sequential patterns enhanced capsule network for sequential recommendation

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-23 DOI: 10.1007/s10489-024-06159-6

Liang Shunpan, Zheng Zhizhong, Zhang Guozheng, Kong Qianjin

{"title":"SPECN:sequential patterns enhanced capsule network for sequential recommendation","authors":"Liang Shunpan, Zheng Zhizhong, Zhang Guozheng, Kong Qianjin","doi":"10.1007/s10489-024-06159-6","DOIUrl":"10.1007/s10489-024-06159-6","url":null,"abstract":"<div><p>Sequential patterns and the order of items in sequences are particularly important for sequential recommendation (SR), which decide what item will interact with the user. However, there are still some problems for the existing methods: (1) They treat all the features of each item equally, we believe that those features the user pays more attention to of each item play a key role to predict next item. (2) Many methods not only ignore sequential patterns and the order of sequences, but also cannot highlight more important features, they only focus on whether the features exist. To address these issues, we propose the <i>Sequential Patterns Enhanced Capsule Network</i> (SPECN). SPECN leverages a self-attention mechanism, using user information as a guide to highlight the most relevant features for each item, then concatenates these features with the original item features in the sequence.SPECN applies horizontal and vertical capsule networks which package neurons into vectors to extract sequential patterns features and the order of sequences. The horizontal capsule network enhances sequential pattern features by learning both the original and user-focused features of individual or adjacent items, containing original features and those features that the user pays more attention to of single item or adjacent items’ features (features of the previous item that the users pay more attention to and features of the current item) to enhance the sequential patterns features. The vertical capsule network captures finer-grained feature representations for each item, improving the recommendation quality. We conduct several experiments on three real-world datasets to demonstrate the superiority of SPECN, outperforming existing methods in terms of accuracy and robustness.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142870466","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Leveraging meta-data of code for adapting prompt tuning for code summarization

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-23 DOI: 10.1007/s10489-024-06197-0

Zhihua Jiang, Di Wang, Dongning Rao

{"title":"Leveraging meta-data of code for adapting prompt tuning for code summarization","authors":"Zhihua Jiang, Di Wang, Dongning Rao","doi":"10.1007/s10489-024-06197-0","DOIUrl":"10.1007/s10489-024-06197-0","url":null,"abstract":"<div><p>Prompt tuning alleviates the gap between pre-training and fine-tuning and achieves promising results in various natural language processing (NLP) tasks. However, it is nontrivial for adapting prompt tuning in intelligent code tasks since code-specific knowledge such as abstract syntax tree is usually hierarchy-structured and therefore is hard to be converted into plain text. Recent works (e.g., PT4Code) introduce simple task prompts along with a programming language indicator into prompt template, achieving improvement over non-prompting state-of-the-art code models (e.g., CodeT5). Inspired by this, we propose a novel code-specific prompt paradigm, meta-data prompt, which introduces semi-structured code’s meta-data (attribute-value pairs) into prompt template and facilitates the adaption of prompt tuning techniques into code tasks. Specifically, we find the usage of diverse meta-data attributes and their combinations and employ the OpenPrompt to implement a meta-data prompt based code model, <b>PRIME</b> (<b>PR</b>ompt tun<b>I</b>ng with <b>ME</b>ta-data), via utilizing CodeT5 as the backbone model. We experiment PRIME with the source code summarization task on the publicly available CodeSearchNet benchmark. Results show that 1) using good meta-data can lead to an improvement on the model performance; 2) the proposed meta-data prompt can be combined with traditional task prompt for further improvement; 3) our best-performing model can consistently outperform CodeT5 by an absolute score of 0.73 and PT4Code by an absolute score of 0.48 regarding the averaged BLEU metric across six programming languages.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 2","pages":""},"PeriodicalIF":3.4,"publicationDate":"2024-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142875229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Towards more accurate object detection via encoding reinforcement and multi-channel enhancement

IF 3.4 2区计算机科学

Applied Intelligence Pub Date : 2024-12-23 DOI: 10.1007/s10489-024-06200-8

Weina Wang, Shuangyong Li, Huxidan Jumahong

引用次数: 0