Yi Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Yi Zhang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Zhuang Wu
School of Management and Engineering Capital University of Economics and Business 100070 Beijing China
5 shared publications - Yuanyuan Wang
School of Management Engineering, Capital University of Economics and Business, 100070 Beijing, China
4 shared publications - Jiaqi Du
School of Management Engineering, Capital University of Economics and Business, 100070 Beijing, China
3 shared publications - Wanshu Fu
School of Management Engineering, Capital University of Economics and Business, 100070 Beijing, China
3 shared publications - Chenjun Liu
School of Management Engineering, Capital University of Economics and Business, 100070 Beijing, China
2 shared publications - Ziqi Wang
Faculty of Engineering and Information Technology, The University of Melbourne, 1446535 Melbourne, Australia
2 shared publications
Latest Publications
- Open Access
- 30 September 2024
- Published by Acadlore Publishing Services Limited in Acadlore Transactions on AI and Machine Learning
- Vol. 3 (3) , 172-182
Predicting Bank Users’ Time Deposits Based on LSTM-Stacked Modeling
Accurately predicting whether bank users will opt for time deposit products is critical for optimizing marketing strategies and enhancing user engagement, ultimately improving a bank’s profitability. Traditional predictive models, such as linear regression and Logistic Regression (LR), are often limited in their ability to capture the complex, time-dependent patterns in user behavior. In this study, a hybrid approach that combines Long Short-Term Memory (LSTM) neural networks and a stacked ensemble learning framework is proposed to address these limitations. Initially, LSTM models were employed to extract temporal features from two distinct bank marketing datasets, thereby capturing the sequential nature of user interactions. These extracted features were subsequently input into several base classifiers, including Random Forest (RF), Support Vector Machine (SVM), and k-Nearest Neighbour (KNN), to conduct initial classifications. The outputs of these classifiers were then integrated using a LR model for final decision-making through a stacking ensemble method. The experimental evaluation demonstrates that the proposed LSTM-stacked model outperforms traditional models in predicting user time deposits on both datasets, providing robust predictive performance. The results suggest that leveraging temporal feature extraction with LSTM and combining it with ensemble techniques yields superior prediction accuracy, thereby offering a more sophisticated solution for banks aiming to enhance their marketing efficiency.
- 30 September 2024
- Published by Acadlore Publishing Services Limited in Acadlore Transactions on AI and Machine Learning
- Vol. 3 (3) , 172-182
Open AccessPredicting Bank Users’ Time Deposits Based on LSTM-Stacked Modeling
Accurately predicting whether bank users will opt for time deposit products is critical for optimizing marketing strategies and enhancing user engagement, ultimately improving a bank’s profitability. Traditional predictive models, such as linear regression and Logistic Regression (LR), are often limited in their ability to capture the complex, time-dependent patterns in user behavior. In this study, a hybrid approach that combines Long Short-Term Memory (LSTM) neural networks and a stacked ensemble learning framework is proposed to address these limitations. Initially, LSTM models were employed to extract temporal features from two distinct bank marketing datasets, thereby capturing the sequential nature of user interactions. These extracted features were subsequently input into several base classifiers, including Random Forest (RF), Support Vector Machine (SVM), and k-Nearest Neighbour (KNN), to conduct initial classifications. The outputs of these classifiers were then integrated using a LR model for final decision-making through a stacking ensemble method. The experimental evaluation demonstrates that the proposed LSTM-stacked model outperforms traditional models in predicting user time deposits on both datasets, providing robust predictive performance. The results suggest that leveraging temporal feature extraction with LSTM and combining it with ensemble techniques yields superior prediction accuracy, thereby offering a more sophisticated solution for banks aiming to enhance their marketing efficiency.
- Open Access
- 14 June 2024
- Published by Acadlore Publishing Services Limited in Information Dynamics and Applications
- Vol. 3 (2) , 89-103
Advancements in Image Recognition: A Siamese Network Approach
In the realm of computer vision, image recognition serves as a pivotal task with extensive applications in intelligent security, autonomous driving, and robotics. Traditional methodologies for image recognition often grapple with computational inefficiencies and diminished accuracy in complex scenarios and extensive datasets. To address these challenges, an algorithm utilizing a siamese network architecture has been developed. This architecture leverages dual interconnected neural network submodules for the efficient extraction and comparison of image features. The effectiveness of this siamese network-based algorithm is demonstrated through its application to various benchmark datasets, where it consistently outperforms conventional approaches in terms of accuracy and processing speed. By employing weight-sharing techniques and optimizing neural network pathways, the proposed algorithm enhances the robustness and efficiency of image recognition tasks. The advancements presented in this study not only contribute to the theoretical understanding but also offer practical solutions, underscoring the significant potential and applicability of siamese networks in advancing image recognition technologies.
- 14 June 2024
- Published by Acadlore Publishing Services Limited in Information Dynamics and Applications
- Vol. 3 (2) , 89-103
Open AccessAdvancements in Image Recognition: A Siamese Network Approach
In the realm of computer vision, image recognition serves as a pivotal task with extensive applications in intelligent security, autonomous driving, and robotics. Traditional methodologies for image recognition often grapple with computational inefficiencies and diminished accuracy in complex scenarios and extensive datasets. To address these challenges, an algorithm utilizing a siamese network architecture has been developed. This architecture leverages dual interconnected neural network submodules for the efficient extraction and comparison of image features. The effectiveness of this siamese network-based algorithm is demonstrated through its application to various benchmark datasets, where it consistently outperforms conventional approaches in terms of accuracy and processing speed. By employing weight-sharing techniques and optimizing neural network pathways, the proposed algorithm enhances the robustness and efficiency of image recognition tasks. The advancements presented in this study not only contribute to the theoretical understanding but also offer practical solutions, underscoring the significant potential and applicability of siamese networks in advancing image recognition technologies.
- Open Access
- 15 April 2024
- Published by Acadlore Publishing Services Limited in International Journal of Knowledge and Innovation Studies
- Vol. 2 (2) , 70-80
A Blockchain Cross-Chain Solution Based on Relays
Blockchain has attracted widespread attention due to its unique features such as decentralization, traceability, and tamper resistance. With the rapid development of blockchain technology, an increasing number of industries are gradually applying blockchain technology to various fields such as the Internet of Things, healthcare, finance, agriculture, and government affairs. However, there are certain differences in the underlying architecture, data structures, consensus algorithms, and other aspects of blockchain technology across different sectors, which restrict transactions to occur within a single blockchain. Achieving interoperability between different blockchains is challenging, hindering data exchange and collaborative business to some extent, inevitably leading to the problem of “data silo”. Against this backdrop, this study aims to explore a cross-chain solution based on relay technology to address the current challenges of interoperability between blockchain systems. By employing relay-based cross-chain technology, a blockchain cross-chain collaboration platform is established to simulate the construction of a real cross-chain network. By deploying business contracts, data and resources between heterogeneous blockchains can seamlessly communicate, resolving the challenge of cross-chain interoperability. The research findings demonstrate that the blockchain cross-chain solution based on relay technology can effectively enhance interoperability between different blockchain systems, enabling cross-chain asset circulation and information transmission, highlighting the practical applicability and scalability of this study.
- 15 April 2024
- Published by Acadlore Publishing Services Limited in International Journal of Knowledge and Innovation Studies
- Vol. 2 (2) , 70-80
Open AccessA Blockchain Cross-Chain Solution Based on Relays
Blockchain has attracted widespread attention due to its unique features such as decentralization, traceability, and tamper resistance. With the rapid development of blockchain technology, an increasing number of industries are gradually applying blockchain technology to various fields such as the Internet of Things, healthcare, finance, agriculture, and government affairs. However, there are certain differences in the underlying architecture, data structures, consensus algorithms, and other aspects of blockchain technology across different sectors, which restrict transactions to occur within a single blockchain. Achieving interoperability between different blockchains is challenging, hindering data exchange and collaborative business to some extent, inevitably leading to the problem of “data silo”. Against this backdrop, this study aims to explore a cross-chain solution based on relay technology to address the current challenges of interoperability between blockchain systems. By employing relay-based cross-chain technology, a blockchain cross-chain collaboration platform is established to simulate the construction of a real cross-chain network. By deploying business contracts, data and resources between heterogeneous blockchains can seamlessly communicate, resolving the challenge of cross-chain interoperability. The research findings demonstrate that the blockchain cross-chain solution based on relay technology can effectively enhance interoperability between different blockchain systems, enabling cross-chain asset circulation and information transmission, highlighting the practical applicability and scalability of this study.
- Open Access
- 31 March 2024
- Published by Acadlore Publishing Services Limited in Journal of Industrial Intelligence
- Vol. 2 (1) , 31-41
An Advanced YOLOv5s Approach for Vehicle Detection Integrating Swin Transformer and SimAM in Dense Traffic Surveillance
In the realm of high-definition surveillance for dense traffic environments, the accurate detection and classification of vehicles remain paramount challenges, often hindered by missed detections and inaccuracies in vehicle type identification. Addressing these issues, an enhanced version of the You Only Look Once version v5s (YOLOv5s) algorithm is presented, wherein the conventional network structure is optimally modified through the partial integration of the Swin Transformer V2. This innovative approach leverages the convolutional neural networks' (CNNs) proficiency in local feature extraction alongside the Swin Transformer V2's capability in global representation capture, thereby creating a symbiotic system for improved vehicle detection. Furthermore, the introduction of the Similarity-based Attention Module (SimAM) within the CNN framework plays a pivotal role, dynamically refocusing the feature map to accentuate local features critical for accurate detection. An empirical evaluation of this augmented YOLOv5s algorithm demonstrates a significant uplift in performance metrics, evidencing an average detection precision (mAP@0.5:0.95) of 65.7%. Specifically, in the domain of vehicle category identification, a notable increase in the true positive rate by 4.48% is observed, alongside a reduction in the false negative rate by 4.11%. The culmination of these enhancements through the integration of Swin Transformer and SimAM within the YOLOv5s framework marks a substantial advancement in the precision of vehicle type recognition and reduction of target miss detection in densely populated traffic flows. The methodology's success underscores the efficacy of this integrated approach in overcoming the prevalent limitations of existing vehicle detection algorithms under complex surveillance scenarios.
- 31 March 2024
- Published by Acadlore Publishing Services Limited in Journal of Industrial Intelligence
- Vol. 2 (1) , 31-41
Open AccessAn Advanced YOLOv5s Approach for Vehicle Detection Integrating Swin Transformer and SimAM in Dense Traffic Surveillance
In the realm of high-definition surveillance for dense traffic environments, the accurate detection and classification of vehicles remain paramount challenges, often hindered by missed detections and inaccuracies in vehicle type identification. Addressing these issues, an enhanced version of the You Only Look Once version v5s (YOLOv5s) algorithm is presented, wherein the conventional network structure is optimally modified through the partial integration of the Swin Transformer V2. This innovative approach leverages the convolutional neural networks' (CNNs) proficiency in local feature extraction alongside the Swin Transformer V2's capability in global representation capture, thereby creating a symbiotic system for improved vehicle detection. Furthermore, the introduction of the Similarity-based Attention Module (SimAM) within the CNN framework plays a pivotal role, dynamically refocusing the feature map to accentuate local features critical for accurate detection. An empirical evaluation of this augmented YOLOv5s algorithm demonstrates a significant uplift in performance metrics, evidencing an average detection precision (mAP@0.5:0.95) of 65.7%. Specifically, in the domain of vehicle category identification, a notable increase in the true positive rate by 4.48% is observed, alongside a reduction in the false negative rate by 4.11%. The culmination of these enhancements through the integration of Swin Transformer and SimAM within the YOLOv5s framework marks a substantial advancement in the precision of vehicle type recognition and reduction of target miss detection in densely populated traffic flows. The methodology's success underscores the efficacy of this integrated approach in overcoming the prevalent limitations of existing vehicle detection algorithms under complex surveillance scenarios.
- Open Access
- 28 November 2023
- Published by Acadlore Publishing Services Limited in Opportunities and Challenges in Sustainability
- Vol. 2 (4) , 184-196
Spatial and Seasonal Dynamics of Air Quality in the Beijing-Tianjin-Hebei Region: An Analysis Using K-means Clustering and BP Neural Networks
In the context of rapid economic development, air pollution has emerged as a critical environmental issue, particularly in the Beijing-Tianjin-Hebei region. This study, through the application of Air Quality Index (AQI) data and K-means clustering, investigates the seasonal variations and spatial distribution of air quality in this region. It has been identified that air pollution in this area is not only subject to seasonal fluctuation but also exhibits distinct patterns of local spatial aggregation. Utilizing a Back Propagation (BP) Neural Network model, this research predicts AQI values, offering foresight into the development and transformation of haze weather conditions. The findings of this investigation are instrumental in enhancing the understanding of air pollution dynamics, facilitating the formulation of effective air control strategies. Such strategies are vital for the issuance of accurate pollution warnings and reminders, thereby contributing to the mitigation of severe pollution impacts.
- 28 November 2023
- Published by Acadlore Publishing Services Limited in Opportunities and Challenges in Sustainability
- Vol. 2 (4) , 184-196
Open AccessSpatial and Seasonal Dynamics of Air Quality in the Beijing-Tianjin-Hebei Region: An Analysis Using K-means Clustering and BP Neural Networks
In the context of rapid economic development, air pollution has emerged as a critical environmental issue, particularly in the Beijing-Tianjin-Hebei region. This study, through the application of Air Quality Index (AQI) data and K-means clustering, investigates the seasonal variations and spatial distribution of air quality in this region. It has been identified that air pollution in this area is not only subject to seasonal fluctuation but also exhibits distinct patterns of local spatial aggregation. Utilizing a Back Propagation (BP) Neural Network model, this research predicts AQI values, offering foresight into the development and transformation of haze weather conditions. The findings of this investigation are instrumental in enhancing the understanding of air pollution dynamics, facilitating the formulation of effective air control strategies. Such strategies are vital for the issuance of accurate pollution warnings and reminders, thereby contributing to the mitigation of severe pollution impacts.
- Open Access
- 30 September 2023
- Published by Acadlore Publishing Services Limited in Journal of Urban Development and Management
- Vol. 2 (3) , 145-159
Spatio-Temporal Dynamics and Haze Agglomeration Analysis in the Beijing-Tianjin-Hebei Region: A WOA-LSTM Approach
The intensification of industrial and urban growth has precipitated a significant increase in atmospheric pollutant emissions, thereby exacerbating air quality deterioration. This phenomenon is particularly pronounced within the Beijing-Tianjin-Hebei urban agglomeration, where haze events have manifested with increasing frequency. Prior investigations have predominantly concentrated on temporal trends, often overlooking the critical impact of geographical factors on haze development. This research delves into the spatio-temporal distribution traits of haze within the Beijing-Tianjin-Hebei region, employing a Whale Optimization Algorithm-Long Short-Term Memory (WOA-LSTM) model. Findings indicate a pronounced spatial concentration of urban air pollution in the region's southern sector. In terms of temporal distribution, the Air Quality Index (AQI) demonstrates distinct seasonal fluctuations, with the highest pollution levels recorded in winter and notably lower levels observed during summer. The study's innovation lies in the development of a WOA-LSTM model, which not only predicts the AQI-a comprehensive haze pollution index-but also offers early warnings pertinent to public travel. By integrating extensive datasets and applying advanced analytical techniques, the study contributes significantly to understanding the complex interplay between urban dynamics and haze distribution. The research underscores the necessity for regional policies tailored to specific spatiotemporal characteristics, thereby aiding in effective air quality management and mitigation strategies within urban agglomerations.
- 30 September 2023
- Published by Acadlore Publishing Services Limited in Journal of Urban Development and Management
- Vol. 2 (3) , 145-159
Open AccessSpatio-Temporal Dynamics and Haze Agglomeration Analysis in the Beijing-Tianjin-Hebei Region: A WOA-LSTM Approach
The intensification of industrial and urban growth has precipitated a significant increase in atmospheric pollutant emissions, thereby exacerbating air quality deterioration. This phenomenon is particularly pronounced within the Beijing-Tianjin-Hebei urban agglomeration, where haze events have manifested with increasing frequency. Prior investigations have predominantly concentrated on temporal trends, often overlooking the critical impact of geographical factors on haze development. This research delves into the spatio-temporal distribution traits of haze within the Beijing-Tianjin-Hebei region, employing a Whale Optimization Algorithm-Long Short-Term Memory (WOA-LSTM) model. Findings indicate a pronounced spatial concentration of urban air pollution in the region's southern sector. In terms of temporal distribution, the Air Quality Index (AQI) demonstrates distinct seasonal fluctuations, with the highest pollution levels recorded in winter and notably lower levels observed during summer. The study's innovation lies in the development of a WOA-LSTM model, which not only predicts the AQI-a comprehensive haze pollution index-but also offers early warnings pertinent to public travel. By integrating extensive datasets and applying advanced analytical techniques, the study contributes significantly to understanding the complex interplay between urban dynamics and haze distribution. The research underscores the necessity for regional policies tailored to specific spatiotemporal characteristics, thereby aiding in effective air quality management and mitigation strategies within urban agglomerations.
- Open Access
- 31 March 2023
- Published by Acadlore Publishing Services Limited in Journal of Urban Development and Management
- Vol. 2 (1) , 34-46
Optimizing Emergency Supply Location Selection in Urban Areas: A Multi-Objective Planning Model and Algorithm
The scientific location and layout of emergency material storage and rescue points in urban areas are critical aspects of emergency management. In this study, a multi-objective programming optimization model was constructed based on related theories, incorporating multiple goal combinations with different dimensions according to various disaster scenarios and urban emergency needs. The weight factors of emergency timeliness, economy, and safety were considered, and the multi-objective model optimization problem was transformed into a single-objective comprehensive optimization model problem using the weight method. The analysis decision function was utilized to study the transformation and solution method of the urban emergency rescue point location model. Heuristic optimization algorithms were employed to perform average segmentation calculations on the preset neighborhoods, constantly changing and narrowing the neighborhood range until the algorithm termination conditions were met, approaching the domain range of the optimal solution. Additionally, another precision parameter was utilized to control the accuracy of the final solution neighborhood range. The optimization of emergency vehicle scheduling was used to synergistically solve the problem of reserve rescue point location layout and optimization solution. The results of the example demonstrate the feasibility of constructing a multi-objective model with multiple combinations of different dimensions of objectives and the rationality of the Dijkstra heuristic optimization algorithm used. This study provides multiple methodologies and alternative site selection plans for decision-makers to select the required multi-objective reserve rescue point location model based on different urban disaster situations and their own emergency rescue needs.
- 31 March 2023
- Published by Acadlore Publishing Services Limited in Journal of Urban Development and Management
- Vol. 2 (1) , 34-46
Open AccessOptimizing Emergency Supply Location Selection in Urban Areas: A Multi-Objective Planning Model and Algorithm
The scientific location and layout of emergency material storage and rescue points in urban areas are critical aspects of emergency management. In this study, a multi-objective programming optimization model was constructed based on related theories, incorporating multiple goal combinations with different dimensions according to various disaster scenarios and urban emergency needs. The weight factors of emergency timeliness, economy, and safety were considered, and the multi-objective model optimization problem was transformed into a single-objective comprehensive optimization model problem using the weight method. The analysis decision function was utilized to study the transformation and solution method of the urban emergency rescue point location model. Heuristic optimization algorithms were employed to perform average segmentation calculations on the preset neighborhoods, constantly changing and narrowing the neighborhood range until the algorithm termination conditions were met, approaching the domain range of the optimal solution. Additionally, another precision parameter was utilized to control the accuracy of the final solution neighborhood range. The optimization of emergency vehicle scheduling was used to synergistically solve the problem of reserve rescue point location layout and optimization solution. The results of the example demonstrate the feasibility of constructing a multi-objective model with multiple combinations of different dimensions of objectives and the rationality of the Dijkstra heuristic optimization algorithm used. This study provides multiple methodologies and alternative site selection plans for decision-makers to select the required multi-objective reserve rescue point location model based on different urban disaster situations and their own emergency rescue needs.
- Open Access
- 29 November 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (23) , 15922
Research on Optimizing the Location and Layout of National Emergency Material Reserve
Scientific planning, and the layout of the national level reserve base of emergency materials, will help improve the effectiveness of a country’s overall emergency disaster reduction system. Based on the P-center location theory, this paper analyzes the factors affecting site selection at the national and state levels, determines the reasonable number of emergency material reserve bases at national and state levels, and then develops a national macro-level emergency supply reserve location planning model. In this study, the city of 28 states in a country were selected as alternative reserve cities and emergency demand city matrix, to conduct the research. The model is solved using a variable neighborhood search algorithm (VNS). The calculation results obtain the reasonable number of emergency material reserve bases set at the national level in this studied country, and the optimal solution of the base layout can be obtained if the number of emergency reserve bases set at the national level remains unchanged. The experimental results shows that the selected algorithm is reasonable.
- 29 November 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (23) , 15922
Open AccessResearch on Optimizing the Location and Layout of National Emergency Material Reserve
Scientific planning, and the layout of the national level reserve base of emergency materials, will help improve the effectiveness of a country’s overall emergency disaster reduction system. Based on the P-center location theory, this paper analyzes the factors affecting site selection at the national and state levels, determines the reasonable number of emergency material reserve bases at national and state levels, and then develops a national macro-level emergency supply reserve location planning model. In this study, the city of 28 states in a country were selected as alternative reserve cities and emergency demand city matrix, to conduct the research. The model is solved using a variable neighborhood search algorithm (VNS). The calculation results obtain the reasonable number of emergency material reserve bases set at the national level in this studied country, and the optimal solution of the base layout can be obtained if the number of emergency reserve bases set at the national level remains unchanged. The experimental results shows that the selected algorithm is reasonable.
- Open Access
- 7 November 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (21) , 14657
A Multistage Time-Delay Control Model for COVID-19 Transmission
With the transmission of the COVID-19 epidemic at home and abroad, this paper considers the spread process in China, improves the classic epidemic SEIR model, and establishes a multistage time-delay control model (MTCM) for COVID-19 transmission. The MTCM divides the spread of COVID-19 into three periods: the outbreak period, the control period and the steady period. The classical SEIR model, the improved SEQIR model and the SEQIR Ⅱ model correspond to the three periods. The classical SEIR model was adopted for the outbreak period and yielded results that were consistent with the observed early propagation of COVID-19 transmission. In the control period, adding isolation measures and a time delay to the MTCM and adjusting the rates yielded a better simulation effect. In the steady period, the focus of consideration is the number of new patients, population movement (in-migration and out-migration of the population) and patient classification (symptomatic and asymptomatic patients). The MCTM was used for simulation, and the comparison results revealed that the simulated data of the MCTM (improved SEQIR model) and the actual data are similar in the control period. The control policy of isolation measures is effective. New infections, population flow and patients with symptomatic or asymptomatic symptoms are more consistent with the steady period characteristics. The multi-stage time-delay control model for COVID-19 transmission provides theoretical methods and good prevention and control measures for future epidemic policy formulation.
- 7 November 2022
- Published by MDPI AG in Sustainability
- Vol. 14 (21) , 14657
Open AccessA Multistage Time-Delay Control Model for COVID-19 Transmission
With the transmission of the COVID-19 epidemic at home and abroad, this paper considers the spread process in China, improves the classic epidemic SEIR model, and establishes a multistage time-delay control model (MTCM) for COVID-19 transmission. The MTCM divides the spread of COVID-19 into three periods: the outbreak period, the control period and the steady period. The classical SEIR model, the improved SEQIR model and the SEQIR Ⅱ model correspond to the three periods. The classical SEIR model was adopted for the outbreak period and yielded results that were consistent with the observed early propagation of COVID-19 transmission. In the control period, adding isolation measures and a time delay to the MTCM and adjusting the rates yielded a better simulation effect. In the steady period, the focus of consideration is the number of new patients, population movement (in-migration and out-migration of the population) and patient classification (symptomatic and asymptomatic patients). The MCTM was used for simulation, and the comparison results revealed that the simulated data of the MCTM (improved SEQIR model) and the actual data are similar in the control period. The control policy of isolation measures is effective. New infections, population flow and patients with symptomatic or asymptomatic symptoms are more consistent with the steady period characteristics. The multi-stage time-delay control model for COVID-19 transmission provides theoretical methods and good prevention and control measures for future epidemic policy formulation.