Resource Type

Journal Article 261

Year

2023 48

2022 44

2021 37

2020 30

2019 29

2018 21

2017 18

2016 11

2015 1

2011 1

2010 1

2009 1

2008 2

2007 6

2006 1

2005 2

2004 2

2002 2

2001 3

1999 1

open ︾

Keywords

Machine learning 42

Deep learning 34

Artificial intelligence 14

Reinforcement learning 14

Active learning 4

Bayesian optimization 3

Big data 3

Data-driven 3

Adaptive dynamic programming 2

Additive manufacturing 2

Attention 2

Autonomous driving 2

Autonomous learning 2

COVID-19 2

Classification 2

Feature learning 2

Materials design 2

Neural network 2

Representation learning 2

open ︾

Search scope:

排序： Display mode:

Correspondence: Uncertainty-aware complementary label queries for active learning Perspective

Shengyuan LIU, Ke CHEN, Tianlei HU, Yunqing MAO,liushengyuan@zju.edu.cn,chenk@cs.zju.edu.cn,htl@zju.edu.cn,myq@citycloud.com.cn

Frontiers of Information Technology & Electronic Engineering 2023, Volume 24, Issue 10,   Pages 1497-1503 doi: 10.1631/FITEE.2200589

Abstract: Many active learning methods assume that a learner can simply ask for the full annotations of some training data from annotators. These methods mainly try to cut the annotation costs by minimizing the number of annotation actions. Unfortunately, annotating instances exactly in many real-world classification tasks is still expensive. To reduce the cost of a single annotation action, we try to tackle a novel active learning setting, named active learning with complementary labels (ALCL). ALCL learners ask only yes/no questions in some classes. After receiving answers from annotators, ALCL learners obtain a few supervised instances and more training instances with complementary labels, which specify only one of the classes to which the pattern does not belong. There are two challenging issues in ALCL: one is how to sample instances to be queried, and the other is how to learn from these complementary labels and ordinary accurate labels. For the first issue, we propose an uncertainty-based sampling strategy under this novel setup. For the second issue, we upgrade a previous ALCL method to fit our sampling strategy. Experimental results on various datasets demonstrate the superiority of our approaches.

Keywords： 主动学习；图片分类；弱监督学习    

HTML PDF Collect

Interactive visual labelling versus active learning: an experimental comparison Research

Mohammad CHEGIN, Jürgen BERNARD, Jian CUI, Fatemeh CHEGINI, Alexei SOURIN, Keith Keith, Tobias SCHRECK

Frontiers of Information Technology & Electronic Engineering 2020, Volume 21, Issue 4,   Pages 524-535 doi: 10.1631/FITEE.1900549

Abstract: Methods from supervised machine learning allow the classification of new data automatically and are tremendously helpful for data analysis. The quality of supervised maching learning depends not only on the type of algorithm used, but also on the quality of the labelled dataset used to train the classifier. Labelling instances in a training dataset is often done manually relying on selections and annotations by expert analysts, and is often a tedious and time-consuming process. Active learning algorithms can automatically determine a subset of data instances for which labels would provide useful input to the learning process. Interactive visual labelling techniques are a promising alternative, providing effective visual overviews from which an analyst can simultaneously explore data records and select items to a label. By putting the analyst in the loop, higher accuracy can be achieved in the resulting classifier. While initial results of interactive visual labelling techniques are promising in the sense that user labelling can improve supervised learning, many aspects of these techniques are still largely unexplored. This paper presents a study conducted using the mVis tool to compare three interactive visualisations, similarity map, scatterplot matrix (SPLOM), and parallel coordinates, with each other and with active learning for the purpose of labelling a multivariate dataset. The results show that all three interactive visual labelling techniques surpass active learning algorithms in terms of classifier accuracy, and that users subjectively prefer the similarity map over SPLOM and parallel coordinates for labelling. Users also employ different labelling strategies depending on the visualisation used.

Keywords： Interactive visual labelling     Active learning     Visual analytics    

HTML PDF Collect

Ensemble enhanced active learning mixture discriminant analysis model and its application for semi-supervised fault classification Research Article

Weijun WANG, Yun WANG, Jun WANG, Xinyun FANG, Yuchen HE

Frontiers of Information Technology & Electronic Engineering 2022, Volume 23, Issue 12,   Pages 1814-1827 doi: 10.1631/FITEE.2200053

Abstract: As an indispensable part of process monitoring, the performance of relies heavily on the sufficiency of process knowledge. However, data labels are always difficult to acquire because of the limited sampling condition or expensive laboratory analysis, which may lead to deterioration of classification performance. To handle this dilemma, a new strategy is performed in which enhanced is employed to evaluate the value of each unlabeled sample with respect to a specific labeled dataset. Unlabeled samples with large values will serve as supplementary information for the training dataset. In addition, we introduce several reasonable indexes and criteria, and thus human labeling interference is greatly reduced. Finally, the effectiveness of the proposed method is evaluated using a numerical example and the Tennessee Eastman process.

Keywords： Semi-supervised     Active learning     Ensemble learning     Mixture discriminant analysis     Fault classification    

HTML PDF Collect

Automatic traceability link recovery via active learning Research Articles

Tian-bao Du, Guo-hua Shen, Zhi-qiu Huang, Yao-shen Yu, De-xiang Wu,tbdu_312@outlook.com,ghshen@nuaa.edu.cn,zqhuang@nuaa.edu.cn

Frontiers of Information Technology & Electronic Engineering 2020, Volume 21, Issue 8,   Pages 1217-1225 doi: 10.1631/FITEE.1900222

Abstract: (TLR) is an important and costly software task that requires humans establish relationships between source and target artifact sets within the same project. Previous research has proposed to establish traceability links by machine learning approaches. However, current machine learning approaches cannot be well applied to projects without traceability information (links), because training an effective predictive model requires humans label too many traceability links. To save , we propose a new TLR approach based on (AL), which is called the AL-based approach. We evaluate the AL-based approach on seven commonly used traceability datasets and compare it with an information retrieval based approach and a state-of-the-art machine learning approach. The results indicate that the AL-based approach outperforms the other two approaches in terms of F-score.

Keywords： Automatic     Traceability link recovery     Manpower     Active learning    

HTML PDF Collect

Disturbance rejection via iterative learning controlwith a disturbance observer for active magnetic bearing systems None

Ze-zhi TANG, Yuan-jin YU, Zhen-hong LI, Zheng-tao DING

Frontiers of Information Technology & Electronic Engineering 2019, Volume 20, Issue 1,   Pages 131-140 doi: 10.1631/FITEE.1800558

Abstract:

Although standard iterative learning control (ILC) approaches can achieve perfect tracking for active magnetic bearing (AMB) systems under external disturbances, the disturbances are required to be iteration-invariant. In contrast to existing approaches, we address the tracking control problem of AMB systems under iteration-variant disturbances that are in different channels from the control inputs. A disturbance observer based ILC scheme is proposed that consists of a universal extended state observer (ESO) and a classical ILC law. Using only output feedback, the proposed control approach estimates and attenuates the disturbances in every iteration. The convergence of the closed-loop system is guaranteed by analyzing the contraction behavior of the tracking error. Simulation and comparison studies demonstrate the superior tracking performance of the proposed control approach.

Keywords： Active magnetic bearings (AMBs)     Iterative learning control (ILC)     Disturbance observer    

HTML PDF Collect

A deep Q-learning network based active object detection model with a novel training algorithm for service robots Research Article

Shaopeng LIU, Guohui TIAN, Yongcheng CUI, Xuyang SHAO

Frontiers of Information Technology & Electronic Engineering 2022, Volume 23, Issue 11,   Pages 1673-1683 doi: 10.1631/FITEE.2200109

Abstract:

This paper focuses on the problem of (AOD). AOD is important for to complete tasks in the family environment, and leads robots to approach the target object by taking appropriate moving actions. Most of the current AOD methods are based on reinforcement learning with low training efficiency and testing accuracy. Therefore, an AOD model based on a (DQN) with a novel training algorithm is proposed in this paper. The DQN model is designed to fit the Q-values of various actions, and includes state space, feature extraction, and a multilayer perceptron. In contrast to existing research, a novel training algorithm based on memory is designed for the proposed DQN model to improve training efficiency and testing accuracy. In addition, a method of generating the end state is presented to judge when to stop the AOD task during the training process. Sufficient comparison experiments and ablation studies are performed based on an AOD dataset, proving that the presented method has better performance than the comparable methods and that the proposed training algorithm is more effective than the raw training algorithm.

Keywords： Active object detection     Deep Q-learning network     Training method     Service robots    

HTML PDF Collect

TC Assurance Architecture for Cybersecurity Infrastructure Based on Active Defense

Zhang Dawei,Shen Changxiang, Liu Jiqiang, Zhang Feifei, Li Lun, Cheng Lichen

Strategic Study of CAE 2016, Volume 18, Issue 6,   Pages 58-61 doi: 10.15302/J-SSCAE-2016.06.012

Abstract:

This paper introduces the status, problems, and future strategies of the cyberspace security infrastructure system, and proposes that cyberspace security infrastructure must be based on active defense. Therefore, this paper proposes several suggestions for a trusted technology insurance system, which include the following: In order to build a trusted technology insurance system, independent innovation in active defense must be the breaking point; key information security systems must be developed by local institutions; independent innovation must be increased; research, product development, and active defense applications must be promoted; the development of trusted computing standards must be promoted; and experimental demonstrations must be carried out.

Keywords： active defense     active immunity     trusted computing     trusted technology insurance system     cybersecurity infrastructure    

HTML PDF Collect

Model and algorithm for satellite’s active temperature control loop design

Li Yunze,Yang Juan,Ning Xianwen,Wang Xiaoming,Shi Xiaobo

Strategic Study of CAE 2008, Volume 10, Issue 7,   Pages 48-50

Abstract:

Electrical heater is an important type of active thermal control methods. A dynamical equation of its control object's transient temperature change is established through principle analysis. The heating power demand calculation models at different temperature control models have been developed from the dynamical equation. The calculating flow of electrical heater's heat power has also been introduced. An active heating loop design example, which is included at the end of this paper, shows that the equations developed above can be used conveniently in satellite's active thermal control system designing and calculating.

Keywords： satellite     active thermal control     active temperature control loop     model and algorithm    

HTML PDF Collect

Research on Active Handoff Mechanism in Micro-Mobility Protocols

Zhao Aqun

Strategic Study of CAE 2004, Volume 6, Issue 8,   Pages 50-56

Abstract:

Aiming at the disadvantages of current handoff mechanisms in micro-mobility protocols, active handoff mechanism is proposed. This mechanism takes advantage of mobility prediction technique to predict the next cell a mobile host will handoff to and the handoff time before the handoff really happens and establish the new path for the mobile host in advance. To guarantee the implementation of active handoff mechanism, a mobility prediction algorithm is proposed which is suitable for active handoff mechanism and easy to implement. Meanwhile the methods to avoid packet loss and packet duplication in active handoff procedure are proposed. The performance of active handoff mechanism is evaluated through theoretical analysis and system simulation. The result shows that active handoff mechanism obtains considerable performance improvement with less cost.

Keywords： micro-mobility protocols     active handoff mechanism     mobility prediction     performance evaluation    

HTML PDF Collect

Active Machine Learning for Chemical Engineers: A Bright Future Lies Ahead! Perspective

Yannick Ureel, Maarten R. Dobbelaere, Yi Ouyang, Kevin De Ras, Maarten K. Sabbe, Guy B. Marin, Kevin M. Van Geem

Engineering 2023, Volume 27, Issue 8,   Pages 23-30 doi: 10.1016/j.eng.2023.02.019

Abstract:

By combining machine learning with the design of experiments, thereby achieving so-called active machine learning, more efficient and cheaper research can be conducted. Machine learning algorithms are more flexible and are better than traditional design of experiment algorithms at investigating processes spanning all length scales of chemical engineering. While active machine learning algorithms are maturing, their applications are falling behind. In this article, three types of challenges presented by active machine learning—namely, convincing the experimental researcher, the flexibility of data creation, and the robustness of active machine learning algorithms—are identified, and ways to overcome them are discussed. A bright future lies ahead for active machine learning in chemical engineering, thanks to increasing automation and more efficient algorithms that can drive novel discoveries. 

Keywords： Active machine learning     Active learning     Bayesian optimization     Chemical engineering     Design of experiments    

HTML PDF Collect

Framework and Application Strategy of Smart Proactive Health Service

Liang Chao , Wang Hua , Tang Lixu

Strategic Study of CAE 2023, Volume 25, Issue 5,   Pages 30-42 doi: 10.15302/J-SSCAE-2023.05.014

Abstract:
The innovative application of smart proactive health services is an important component of the Healthy China initiative and an effective measure to satisfy the diversified health needs of the public in the post-epidemic era. This study focuses on building a new paradigm of smart proactive health services and aims to improve proactive health intervention and management capabilities and provide high-quality health services for the public. It analyzes the current status and challenges of proactive health services, summarizes the smart development trend, and proposes the concept of smart proactive health services. Moreover, the study constructs a technical system of smart proactive health services using a structured analysis method, establishes an application framework consisting of one center, one portal, and three endpoints, and proposes a construction ecology featuring internal and external collaboration. The application scenarios and practical cases of the smart proactive health services are summarized in terms of technology integration and smart application. Furthermore, we propose the following suggestions to promote the sustainable and high quality development of smart proactive health services in China: (1) strengthening macro policy tools to improve the development environment; (2) improving the public’s digital literacy to reshape the atmosphere for service participation; (3) building a service standards system to improve the internal digital ecology; (4) creating a multiple supply pattern to continuously improve service quality; and (5) accelerating the integration of industry, academia, research, and application to enhance the commercialization of research findings.

Keywords： smart proactive health services     sports for health     technology system     application framework     construction ecology    

HTML PDF Collect

Self-Powered Active Vibration Control: Concept, Modeling, and Testing Article

Jin-Yang Li, Songye Zhu

Engineering 2022, Volume 11, Issue 4,   Pages 126-137 doi: 10.1016/j.eng.2021.03.022

Abstract:

Despite their superior control performance, active vibration control techniques cannot be widely used in some engineering fields because of their substantial power demand in controlling large-scale structures. As an innovative solution to this problem, an unprecedented self-powered active vibration control system was established in this study. The topological design, working mechanism, and power flow of the proposed system are presented herein. The self-powering ability of the system was confirmed based on a detailed power flow analysis of vibration control processes. A self-powered actively controlled actuator was designed and applied to a scaled active vibration isolation table. The feasibility and effectiveness of the innovative system were successfully validated through a series of analytical, numerical, and experimental investigations. The setup and control strategy of the proposed system can be readily extended to diversified active vibration control applications in various engineering fields.

Keywords： Self-powered active vibration control     Energy harvesting     Skyhook control     Power equilibrium     Smart control    

HTML PDF Collect

Product Modular Design Method for Active Recovery

Zhang Chongyuan,Wei Wei,Zhan Yang,Li Rupeng

Strategic Study of CAE 2018, Volume 20, Issue 2,   Pages 42-49 doi: 10.15302/J-SSCAE-2018.02.007

Abstract:

The demand for product recovery performance has been gradually improved with the environmental protection awareness. Based on the traditional modular design method, this paper integrates the active recovery products modular design idea, puts forward the active recovery products modularization criteria and takes the active recovery, internal polymerization degree and external coupling degree as the optimization target to divide modules. In the algorithm section, this paper proposes the clonal multi-objective optimization algorithm. It is based on mutation operation and optimized by removing more crowded antibodies. Finally, we apply the method to the internal combustion engine and compare the method with the unoptimized algorithm.The conclusion proves the superiority of the improved immune algorithm.

Keywords： active recycling     multiobjective optimization     modular partition     improved immune algorithm     green design    

HTML PDF Collect

Active Support of Power System to Energy Transition Editorial

Yusheng Xue

Engineering 2021, Volume 7, Issue 8,   Pages 1035-1036 doi: 10.1016/j.eng.2021.07.002

HTML PDF Collect

Active Control of Ocean Vehicle Multidimension Vibration and Motion

Chen Xiuxiang,Ma Lüzhong,Wu Weiguang,Zhu Wei

Strategic Study of CAE 2007, Volume 9, Issue 8,   Pages 52-56

Abstract:

Active control multi-dimension vibration device based on parallel mechanisms and electromagnetic actuator is used to realize the multidimensional low-frequency vibration damping of the shipping under sail.  With the vibration control device,  structural design, system control model analysis and control algorithm research,  as well as prototyping testing are carried out.  The results of prototyping testing indicate that with the active control multi-dimension vibration damping the multi-dimensional low-frequency vibration active controled of shipping under sail can be achieved.

Keywords： electromagnetic actuator     active control     multi-dimension vibration     parallel mechanisms    

HTML PDF Collect

Title Author Date Type Operation

Correspondence: Uncertainty-aware complementary label queries for active learning

Shengyuan LIU, Ke CHEN, Tianlei HU, Yunqing MAO,liushengyuan@zju.edu.cn,chenk@cs.zju.edu.cn,htl@zju.edu.cn,myq@citycloud.com.cn

Journal Article

Interactive visual labelling versus active learning: an experimental comparison

Mohammad CHEGIN, Jürgen BERNARD, Jian CUI, Fatemeh CHEGINI, Alexei SOURIN, Keith Keith, Tobias SCHRECK

Journal Article

Ensemble enhanced active learning mixture discriminant analysis model and its application for semi-supervised fault classification

Weijun WANG, Yun WANG, Jun WANG, Xinyun FANG, Yuchen HE

Journal Article

Automatic traceability link recovery via active learning

Tian-bao Du, Guo-hua Shen, Zhi-qiu Huang, Yao-shen Yu, De-xiang Wu,tbdu_312@outlook.com,ghshen@nuaa.edu.cn,zqhuang@nuaa.edu.cn

Journal Article

Disturbance rejection via iterative learning controlwith a disturbance observer for active magnetic bearing systems

Ze-zhi TANG, Yuan-jin YU, Zhen-hong LI, Zheng-tao DING

Journal Article

A deep Q-learning network based active object detection model with a novel training algorithm for service robots

Shaopeng LIU, Guohui TIAN, Yongcheng CUI, Xuyang SHAO

Journal Article

TC Assurance Architecture for Cybersecurity Infrastructure Based on Active Defense

Zhang Dawei,Shen Changxiang, Liu Jiqiang, Zhang Feifei, Li Lun, Cheng Lichen

Journal Article

Model and algorithm for satellite’s active temperature control loop design

Li Yunze,Yang Juan,Ning Xianwen,Wang Xiaoming,Shi Xiaobo

Journal Article

Research on Active Handoff Mechanism in Micro-Mobility Protocols

Zhao Aqun

Journal Article

Active Machine Learning for Chemical Engineers: A Bright Future Lies Ahead!

Yannick Ureel, Maarten R. Dobbelaere, Yi Ouyang, Kevin De Ras, Maarten K. Sabbe, Guy B. Marin, Kevin M. Van Geem

Journal Article

Framework and Application Strategy of Smart Proactive Health Service

Liang Chao , Wang Hua , Tang Lixu

Journal Article

Self-Powered Active Vibration Control: Concept, Modeling, and Testing

Jin-Yang Li, Songye Zhu

Journal Article

Product Modular Design Method for Active Recovery

Zhang Chongyuan,Wei Wei,Zhan Yang,Li Rupeng

Journal Article

Active Support of Power System to Energy Transition

Yusheng Xue

Journal Article

Active Control of Ocean Vehicle Multidimension Vibration and Motion

Chen Xiuxiang,Ma Lüzhong,Wu Weiguang,Zhu Wei

Journal Article