Pub Date : 2026-03-31DOI: 10.1007/s40747-026-02278-4
Yufeng Zhou, Zimei Pan, Xianfang Zeng, Tao Zhou
The highly complex and dynamically uncertain nature of forest fire management necessitates the optimization of task scheduling and emergency resource delivery decisions. To enhance the performance of forest fire emergency response and reduce disaster losses, this study constructs a mixed-integer nonlinear programming model aimed at maximizing firefighting performance in affected areas. The model integrates key factors such as fire spread rate, disaster relief time constraints, and resource demand urgency. It addresses uncertain parameters—such as resource delivery time and fire point recovery time—by applying an interval number-based deterministic processing method. To solve this problem, an improved Tabu Search-Simulated Annealing hybrid algorithm (ITS-SAA), is developed. ITS-SAA improves four types of neighborhood operators and repair operators to improve the algorithm. Compared with TS, SAA, Immune Optimization Algorithm (IOA) and Differential Evolution Algorithm (DE), the ITS-SAA achieves an average optimization improvement of 5.89%, 3.14%, 73.20% and 68.94% respectively. The results show: (1) The ITS-SAA demonstrates both reliability and effectiveness. (2) There exists an optimal threshold for resource allocation. (3) In scenarios with insufficient delivery resources, merely increasing the number of firefighting teams yields limited improvements. Decision-makers should properly configure the number of delivery teams and firefighting teams, with priority given to enhancing logistical delivery capacity.
{"title":"Integrated optimization of forest fire task scheduling and emergency resource delivery under uncertain environments","authors":"Yufeng Zhou, Zimei Pan, Xianfang Zeng, Tao Zhou","doi":"10.1007/s40747-026-02278-4","DOIUrl":"https://doi.org/10.1007/s40747-026-02278-4","url":null,"abstract":"The highly complex and dynamically uncertain nature of forest fire management necessitates the optimization of task scheduling and emergency resource delivery decisions. To enhance the performance of forest fire emergency response and reduce disaster losses, this study constructs a mixed-integer nonlinear programming model aimed at maximizing firefighting performance in affected areas. The model integrates key factors such as fire spread rate, disaster relief time constraints, and resource demand urgency. It addresses uncertain parameters—such as resource delivery time and fire point recovery time—by applying an interval number-based deterministic processing method. To solve this problem, an improved Tabu Search-Simulated Annealing hybrid algorithm (ITS-SAA), is developed. ITS-SAA improves four types of neighborhood operators and repair operators to improve the algorithm. Compared with TS, SAA, Immune Optimization Algorithm (IOA) and Differential Evolution Algorithm (DE), the ITS-SAA achieves an average optimization improvement of 5.89%, 3.14%, 73.20% and 68.94% respectively. The results show: (1) The ITS-SAA demonstrates both reliability and effectiveness. (2) There exists an optimal threshold for resource allocation. (3) In scenarios with insufficient delivery resources, merely increasing the number of firefighting teams yields limited improvements. Decision-makers should properly configure the number of delivery teams and firefighting teams, with priority given to enhancing logistical delivery capacity.","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"50 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-03-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147586439","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}
Pub Date : 2026-03-30DOI: 10.1007/s40747-026-02268-6
Sagvan Y. Musa, Zanyar A. Ameen, Wafa Alagal, Baravan A. Asaad
{"title":"A robust methodology for multi-criteria group decision-making: intuitionistic fuzzy N-bipolar soft expert sets in cybersecurity risk assessment for financial institutions","authors":"Sagvan Y. Musa, Zanyar A. Ameen, Wafa Alagal, Baravan A. Asaad","doi":"10.1007/s40747-026-02268-6","DOIUrl":"https://doi.org/10.1007/s40747-026-02268-6","url":null,"abstract":"","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"17 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147536464","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}
Pub Date : 2026-03-27DOI: 10.1007/s40747-026-02251-1
Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam
Overfitting remains a significant challenge in deep learning, often arising from data outliers, noise, and limited training data. To address this, the Divide2Conquer (D2C) method was previously proposed, utilizing data partitioning as a structural regularizer . By training identical models independently on isolated data shards, this strategy enables learning more consistent patterns while minimizing the influence of global noise. However, D2C’s standard aggregation typically treats all subset models equally, failing to filter out edge models that have overfitted to local noise. Building upon this foundation, we introduce Dynamic Uncertainty-Aware Divide2Conquer (DUA-D2C) , an advanced technique that refines the aggregation process. DUA-D2C dynamically weights the contributions of subset models based on their performance on a shared validation set, employing a novel composite score of accuracy and normalized prediction entropy. This intelligent aggregation allows the central model to preferentially learn from subsets yielding more generalizable and confident edge models. In this work, we provide a rigorous theoretical justification for this approach, analytically demonstrating how dynamic parameter fusion reduces model variance. Empirical evaluations on benchmark datasets spanning image, audio, and text domains demonstrate that DUA-D2C significantly improves generalization. Our analysis includes evaluations of decision boundaries, loss curves, and ablation studies, highlighting that DUA-D2C provides additive performance gains even when applied on top of standard regularizers like Dropout. This study establishes DUA-D2C as a theoretically grounded and effective approach to combating overfitting in modern deep learning. The source codes for this study are available on GitHub at https://github.com/Saiful185/DUAD2C .
{"title":"DUA-D2C: Dynamic uncertainty aware method for overfitting remediation in deep learning","authors":"Md. Saiful Bari Siddiqui, Md Mohaiminul Islam, Md. Golam Rabiul Alam","doi":"10.1007/s40747-026-02251-1","DOIUrl":"https://doi.org/10.1007/s40747-026-02251-1","url":null,"abstract":"Overfitting remains a significant challenge in deep learning, often arising from data outliers, noise, and limited training data. To address this, the Divide2Conquer (D2C) method was previously proposed, utilizing data partitioning as a structural regularizer . By training identical models independently on isolated data shards, this strategy enables learning more consistent patterns while minimizing the influence of global noise. However, D2C’s standard aggregation typically treats all subset models equally, failing to filter out edge models that have overfitted to local noise. Building upon this foundation, we introduce Dynamic Uncertainty-Aware Divide2Conquer (DUA-D2C) , an advanced technique that refines the aggregation process. DUA-D2C dynamically weights the contributions of subset models based on their performance on a shared validation set, employing a novel composite score of accuracy and normalized prediction entropy. This intelligent aggregation allows the central model to preferentially learn from subsets yielding more generalizable and confident edge models. In this work, we provide a rigorous theoretical justification for this approach, analytically demonstrating how dynamic parameter fusion reduces model variance. Empirical evaluations on benchmark datasets spanning image, audio, and text domains demonstrate that DUA-D2C significantly improves generalization. Our analysis includes evaluations of decision boundaries, loss curves, and ablation studies, highlighting that DUA-D2C provides additive performance gains even when applied on top of standard regularizers like Dropout. This study establishes DUA-D2C as a theoretically grounded and effective approach to combating overfitting in modern deep learning. The source codes for this study are available on GitHub at <jats:ext-link xmlns:xlink=\"http://www.w3.org/1999/xlink\" xlink:href=\"https://github.com/Saiful185/DUAD2C\" ext-link-type=\"uri\">https://github.com/Saiful185/DUAD2C</jats:ext-link> .","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"18 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147536046","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}
Pub Date : 2026-03-24DOI: 10.1007/s40747-026-02231-5
Teng Fei, Qin Xin, Jiaming Deng, Jieming Gao, Ge Jin
{"title":"End-to-end high-fidelity avatar reconstruction using 3D Gaussian splatting with monocular video as the sole input","authors":"Teng Fei, Qin Xin, Jiaming Deng, Jieming Gao, Ge Jin","doi":"10.1007/s40747-026-02231-5","DOIUrl":"https://doi.org/10.1007/s40747-026-02231-5","url":null,"abstract":"","PeriodicalId":10524,"journal":{"name":"Complex & Intelligent Systems","volume":"28 1","pages":""},"PeriodicalIF":5.8,"publicationDate":"2026-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147536049","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：