The Dynamic Multi-Objective Bin-Packing Problem (DMOBPP) presents significant challenges in simultaneously optimizing bin usage (z1), thermal heat (z2), and delivery time (z3). This paper proposes AMSS-PATH, a hybrid metaheuristic integrating Strategic Path Relinking into an Adaptive Multi-Objective Scatter Search (AMSS) framework with three specialized operators (Φ1,Φ2,Φ3), each targeting a distinct objective. By exploring elite-solution trajectories, the method densifies the Pareto frontier through targeted intensification at every intermediate point. Experimental results on toy and full-scale instances yield a Net Front Contribution (NFC) of 100%, a Binary Conflict Index (BCI) of ≈ 0, and a 4.8% hypervolume gain over base AMSS, significantly outperforming GAMMA-PC, MOMA, and NSGA-II. A Wilcoxon signed-rank test (p ≈ 0.031) confirms statistical significance.
Sabrin Boulebene, M. Hifi· International Conference on...· 0 citations
The k-Clustering Minimum Biclique Completion Problem (k-CMBCP) is an NP-hard combinatorial optimization problem with significant implications in bipartite graph clustering applications. The objective is to partition a set of services into k disjoint clusters such that the number of missing edges required to transform each cluster into a complete biclique is minimized. This paper proposes a robust Cooperative Scatter Search (CSS) algorithm designed to exploit the structural characteristics of the problem. The proposed metaheuristic framework integrates a diversification-based population initialization strategy, dynamic management of elite reference set, and a specialized uniform-based recombination operator. Furthermore, the algorithm incorporates a sequential Variable Neighborhood Descent (VND) strategy leveraging three complementary neighborhoods to intensify the local search. The contribution is a problem-tailored integration of known search ingredients rather than a generic new metaheuristic paradigm: its novelty lies in the way diversification, reference-set memory, recombination, and SeqVND are coordinated for the specific structure of the k-CMBCP. Computational experiments on benchmark instances show that the proposed method achieves competitive and superior performance compared with recent heuristics in terms of solution quality and computational stability.
M. Hifi, Y. Salmi, Juntao Zhao· International Conference on...· 0 citations
Artificial intelligence holds significant promise for enhancing diagnostic accuracy and mitigating subjectivity in medical imaging, particularly for high-stakes tasks such as cancer detection. However, model performance frequently degrades when transitioned to real-world clinical scenarios. Employing a proprietary clinical dataset consisting of white-light cystoscopy videos acquired from the Amiens University Hospital, this study investigates the design of robust Convolutional Neural Networks (CNNs) through the Non-dominated Sorting Genetic Algorithm (NSGA-II). The search process was configured to automatically generate architectures that maximize both mean sensitivity and mean specificity within a 5-fold cross-validation framework. Search efficiency was evaluated using the HI. The evolved architectures were subsequently benchmarked against VGG16, ResNet50, and DenseNet121 models pretrained on ImageNet. Experimental results show a gradual increase in the HI from 0.60 to 0.67 during the initial phase, followed by a decline upon the inclusion of the third fold, suggesting a "generalization shock" before a final marginal recovery. These fluctuations coincided with a substantial reduction in genotypic diversity from 0.87 to 0.28, leading to premature convergence in the NSGA-II optimization. Consequently, the resulting CNN architecture remained below the ResNet50 baseline, highlighting that further refinement of optimization strategies and training hyperparameters is essential to bridge the gap between automated design and expert-crafted models.
Haithem Dahimi, M. Hifi, F. Saint· International Conference on...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
