Many real-world engineering problems need to balance different objectives and can be formatted as multi-objective optimization problem. An effective multi-objective algorithm can achieve a set of optimal solutions that can make a tradeoff between different objectives, which is valuable to further explore and design. In this paper, an improved multi-objective differential evolution algorithm (MOEA/D/DEM) based on a decomposition strategy is proposed to improve the performance of differential evolution algorithm for practical multi-objective nutrition decision problems. Firstly, considering the neighborhood characteristic, a neighbor intimacy factor is designed in the search process for enhancing the diversity of the population, then a new Gaussian mutation strategy with variable step size is proposed to reduce the probability of escaping local optimum area and improve the local search ability. Finally, the proposed algorithm is tested by classic test problems (DTLZ1-7 and WFG1-9) and applied to the multi-objective nutrition decision problems, compared to the other reported multi-objective algorithms, the proposed algorithm has a better search capability and obtained competitive results.

Chest X-ray images play a crucial role in pneumonia diagnosis, with deep transfer learning being a widely adopted method for pneumonia detection. However, effectively handling feature data extracted from deep models without succumbing to the challenges of feature dimensionality remains a formidable task. In response to this complex issue, we propose a novel two-stage deep feature selection (FS) method utilizing the voting differential evolution (VDE) algorithm. In this approach, a dimension adaptive search strategy is meticulously devised to ensure robust feature selection while concurrently reducing the dimension. To expedite the optimization process, we devise a CR adaptive adjustment method to enhance the efficiency of the algorithm. Notably, an important aspect of our approach is the introduction of a novel DE algorithm that integrates a voting mechanism. This synergistic fusion allows a comprehensive analysis of crucial feature relationships to mitigate the risk of algorithmic entrapment in local optima. Additionally, we propose a dynamic feature evaluation function to avert the oversight of feature sets with optimal classification accuracy during later stages of the algorithm, thereby preserving discriminative features. The method is verified on an open Chest X-Ray Images dataset, achieving 99.04% average precision, 98.67% average accuracy, 99.13% average recall, and 19.93% average feature dimension reduction ratio. The experimental findings reveal that the presented method outperform prevailing state-of-the-art algorithms.