Peng Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Peng Li has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Xiufen Ye
College of Intelligent Systems Science and Engineering, Harbin Engineering University, Harbin, China
2 shared publications - Chuanlong Li
College of Intelligent Systems Science and Engineering, Harbin Engineering University, Harbin 150001, China
1 shared publication - Haibo Yang
College of Automation, Harbin Engineering University, Harbin 150001, China
1 shared publication Haini Guo
School of Management, Harbin University of Commerce, No. 1, Xuehai Street, Harbin 150028, China
1 shared publication- Lin Wang
Monash-Airdoc joint research group, Monash University, Melbourne, VIC, Australia
1 shared publication - Lu Yao
School of Management, Harbin University of Commerce, Harbin, China
1 shared publication
Latest Publications
- Open Access
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 51170-51185
Dual Debiased Recommendation Model Based on Stacked Autoencoder and Reward Mechanism for Exposure Bias and Cold Start
Personalized recommendation systems have become essential in e-commerce platforms, offering tailored product suggestions by leveraging historical interaction data. Despite their widespread adoption, these systems face two persistent challenges: exposure bias and the cold-start problem. Exposure bias leads to an over-concentration of recommendations on high-exposure products, while the cold start problem results from a lack of historical data for new users and products. To address these issues, this paper proposes a dual debiasing recommendation model based on the Stacked Autoencoder-enhanced Neural Collaborative Filtering Recommendation Algorithm with Reward Mechanism (SNCF-RM). The stacked autoencoders can deeply explore user and product features to improve the model's representation ability in cold start scenarios. The reward mechanism increases the recommendation opportunities for low-exposure products by introducing exploratory signals, thereby alleviating exposure bias. Experimental results on two MovieLens datasets show that SNCF-RM improves performance by approximately 12% across eight key metrics, including Precision, Accuracy, RMSE, Hit Ratio, Exposure, Coverage, Novelty and Normalized Discounted Cumulative Gain , confirming its effectiveness in mitigating dual biases.
- 9 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 51170-51185
Open AccessDual Debiased Recommendation Model Based on Stacked Autoencoder and Reward Mechanism for Exposure Bias and Cold Start
Personalized recommendation systems have become essential in e-commerce platforms, offering tailored product suggestions by leveraging historical interaction data. Despite their widespread adoption, these systems face two persistent challenges: exposure bias and the cold-start problem. Exposure bias leads to an over-concentration of recommendations on high-exposure products, while the cold start problem results from a lack of historical data for new users and products. To address these issues, this paper proposes a dual debiasing recommendation model based on the Stacked Autoencoder-enhanced Neural Collaborative Filtering Recommendation Algorithm with Reward Mechanism (SNCF-RM). The stacked autoencoders can deeply explore user and product features to improve the model's representation ability in cold start scenarios. The reward mechanism increases the recommendation opportunities for low-exposure products by introducing exploratory signals, thereby alleviating exposure bias. Experimental results on two MovieLens datasets show that SNCF-RM improves performance by approximately 12% across eight key metrics, including Precision, Accuracy, RMSE, Hit Ratio, Exposure, Coverage, Novelty and Normalized Discounted Cumulative Gain , confirming its effectiveness in mitigating dual biases.
- Open Access
- 10 January 2023
- Published by MDPI AG in Behavioral Sciences
- Vol. 13 (1) , 63
The Influence of Herd Mentality on Rating Bias and Popularity Bias: A Bi-Process Debiasing Recommendation Model Based on Matrix Factorization
To reduce the impact of rating bias and popularity bias in recommender system, and make the recommender system reach a balance between recommendation utility and debias effect at the same time, we propose a bi-process debiasing recommendation model based on matrix factorization. Firstly, considering the problem that the user’s ratings are affected by the herd mentality, which leads to a consistency between the rating and the selection of rating items, resulting in the power-law distribution, the k-times parabolic fuzzy distribution was used to fuse the user’s age to redistribute the ratings. Secondly, the loss function is optimized by the continuously increasing flow and popularity of items. Finally, user emotion and item popularity are combined to construct user psychological tendency, which is divided into three levels: strong, medium and weak, and different levels are given different weights. To verify the performance of the model, the experimental results on real datasets show that the model proposed in this paper not only effectively reduces the recommendation bias but also ensures the recommendation utility.
- 10 January 2023
- Published by MDPI AG in Behavioral Sciences
- Vol. 13 (1) , 63
Open AccessThe Influence of Herd Mentality on Rating Bias and Popularity Bias: A Bi-Process Debiasing Recommendation Model Based on Matrix Factorization
To reduce the impact of rating bias and popularity bias in recommender system, and make the recommender system reach a balance between recommendation utility and debias effect at the same time, we propose a bi-process debiasing recommendation model based on matrix factorization. Firstly, considering the problem that the user’s ratings are affected by the herd mentality, which leads to a consistency between the rating and the selection of rating items, resulting in the power-law distribution, the k-times parabolic fuzzy distribution was used to fuse the user’s age to redistribute the ratings. Secondly, the loss function is optimized by the continuously increasing flow and popularity of items. Finally, user emotion and item popularity are combined to construct user psychological tendency, which is divided into three levels: strong, medium and weak, and different levels are given different weights. To verify the performance of the model, the experimental results on real datasets show that the model proposed in this paper not only effectively reduces the recommendation bias but also ensures the recommendation utility.
- Open Access
ULO: An Underwater Light-Weight Object Detector for Edge Computing
Recent studies on underwater object detection have progressed with the development of deep-learning methods. Generally, the model performance increase is accompanied by an increase in computation. However, a significant fraction of remotely operated underwater vehicles (ROVs) and autonomous underwater vehicles (AUVs) operate in environments with limited power and computation resources, making large models inapplicable. In this paper, we propose a fast and compact object detector—namely, the Underwater Light-weight Object detector (ULO)—for several marine products, such as scallops, starfish, echinus, and holothurians. ULO achieves comparable results to YOLO-v3 with less than 7% of its computation. ULO is modified based on the YOLO Nano architecture, and some modern architectures are used to optimize it, such as the Ghost module and decoupled head design in detection. We propose an adaptive pre-processing module for the image degradation problem that is common in underwater images. The module is lightweight and simple to use, and ablation experiments verify its effectiveness. Moreover, ULO Tiny, a lite version of ULO, is proposed to achieve further computation reduction. Furthermore, we optimize the annotations of the URPC2019 dataset, and the modified annotations are more accurate in localization and classification. The refined annotations are available to the public for research use.
Open AccessULO: An Underwater Light-Weight Object Detector for Edge Computing
Recent studies on underwater object detection have progressed with the development of deep-learning methods. Generally, the model performance increase is accompanied by an increase in computation. However, a significant fraction of remotely operated underwater vehicles (ROVs) and autonomous underwater vehicles (AUVs) operate in environments with limited power and computation resources, making large models inapplicable. In this paper, we propose a fast and compact object detector—namely, the Underwater Light-weight Object detector (ULO)—for several marine products, such as scallops, starfish, echinus, and holothurians. ULO achieves comparable results to YOLO-v3 with less than 7% of its computation. ULO is modified based on the YOLO Nano architecture, and some modern architectures are used to optimize it, such as the Ghost module and decoupled head design in detection. We propose an adaptive pre-processing module for the image degradation problem that is common in underwater images. The module is lightweight and simple to use, and ablation experiments verify its effectiveness. Moreover, ULO Tiny, a lite version of ULO, is proposed to achieve further computation reduction. Furthermore, we optimize the annotations of the URPC2019 dataset, and the modified annotations are more accurate in localization and classification. The refined annotations are available to the public for research use.
- Open Access
- 29 May 2019
- Published by MDPI AG in Remote Sensing
- Vol. 11 (11) , 1281
A Gray Scale Correction Method for Side-Scan Sonar Images Based on Retinex
When side-scan sonars collect data, sonar energy attenuation, the residual of time varying gain, beam patterns, angular responses, and sonar altitude variations occur, which lead to an uneven gray level in side-scan sonar images. Therefore, gray scale correction is needed before further processing of side-scan sonar images. In this paper, we introduce the causes of gray distortion in side-scan sonar images and the commonly used optical and side-scan sonar gray scale correction methods. As existing methods cannot effectively correct distortion, we propose a simple, yet effective gray scale correction method for side-scan sonar images based on Retinex given the characteristics of side-scan sonar images. Firstly, we smooth the original image and add a constant as an illumination map. Then, we divide the original image by the illumination map to produce the reflection map. Finally, we perform element-wise multiplication between the reflection map and a constant coefficient to produce the final enhanced image. Two different schemes are used to implement our algorithm. For gray scale correction of side-scan sonar images, the proposed method is more effective than the latest similar methods based on the Retinex theory, and the proposed method is faster. Experiments prove the validity of the proposed method.
- 29 May 2019
- Published by MDPI AG in Remote Sensing
- Vol. 11 (11) , 1281
Open AccessA Gray Scale Correction Method for Side-Scan Sonar Images Based on Retinex
When side-scan sonars collect data, sonar energy attenuation, the residual of time varying gain, beam patterns, angular responses, and sonar altitude variations occur, which lead to an uneven gray level in side-scan sonar images. Therefore, gray scale correction is needed before further processing of side-scan sonar images. In this paper, we introduce the causes of gray distortion in side-scan sonar images and the commonly used optical and side-scan sonar gray scale correction methods. As existing methods cannot effectively correct distortion, we propose a simple, yet effective gray scale correction method for side-scan sonar images based on Retinex given the characteristics of side-scan sonar images. Firstly, we smooth the original image and add a constant as an illumination map. Then, we divide the original image by the illumination map to produce the reflection map. Finally, we perform element-wise multiplication between the reflection map and a constant coefficient to produce the final enhanced image. Two different schemes are used to implement our algorithm. For gray scale correction of side-scan sonar images, the proposed method is more effective than the latest similar methods based on the Retinex theory, and the proposed method is faster. Experiments prove the validity of the proposed method.
- Open Access
- 18 October 2016
- Published by Wiley in Scientific Programming
- Vol. 2016, 1-9
Optimal Scheduling of Doctors Outpatient Departments Based on Patients’ Behavior
The low operational efficiency in the field of medical and health care has become a serious problem in China; the long time that the patients have to wait for is the main phenomenon of difficult medical service. Medical industry is service-oriented and its main purpose is to make profits, so the benefits and competitiveness of a hospital depend on patient satisfaction. This paper makes a survey on a large hospital in Harbin of China and collects relevant data and then uses the prospect theory to analyze patients’ and doctors’ behavioral characteristics and the model of patient satisfaction is established based on fuzzy theory with a triplet. The optimal scheduling of clinic is described as a problem with the rule of first come, first served which maximizes patient satisfaction for the main goal and minimizes operating costs for the secondary goal. And the corresponding mathematical model is established. Finally, a solution method named plant growth simulation algorithm (PGSA) is presented. Then, by means of calculating of the example and comparing with genetic algorithm, the results show that the optimum can be reached; meanwhile the efficiency of the presented algorithm is better than the genetic algorithm.- 18 October 2016
- Published by Wiley in Scientific Programming
- Vol. 2016, 1-9
Open AccessOptimal Scheduling of Doctors Outpatient Departments Based on Patients’ Behavior
The low operational efficiency in the field of medical and health care has become a serious problem in China; the long time that the patients have to wait for is the main phenomenon of difficult medical service. Medical industry is service-oriented and its main purpose is to make profits, so the benefits and competitiveness of a hospital depend on patient satisfaction. This paper makes a survey on a large hospital in Harbin of China and collects relevant data and then uses the prospect theory to analyze patients’ and doctors’ behavioral characteristics and the model of patient satisfaction is established based on fuzzy theory with a triplet. The optimal scheduling of clinic is described as a problem with the rule of first come, first served which maximizes patient satisfaction for the main goal and minimizes operating costs for the secondary goal. And the corresponding mathematical model is established. Finally, a solution method named plant growth simulation algorithm (PGSA) is presented. Then, by means of calculating of the example and comparing with genetic algorithm, the results show that the optimum can be reached; meanwhile the efficiency of the presented algorithm is better than the genetic algorithm.