Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 2 de 2
Filtrar
Mais filtros










Base de dados
Intervalo de ano de publicação
1.
PLoS One ; 19(6): e0302327, 2024.
Artigo em Inglês | MEDLINE | ID: mdl-38843122

RESUMO

In the context of existing adversarial attack schemes based on unsupervised graph contrastive learning, a common issue arises due to the discreteness of graph structures, leading to reduced reliability of structural gradients and consequently resulting in the problem of attacks getting trapped in local optima. An adversarial attack method based on momentum gradient candidates is proposed in this research. Firstly, the gradients obtained by back-propagation are transformed into momentum gradients, and the gradient update is guided by overlaying the previous gradient information in a certain proportion to accelerate convergence speed and improve the accuracy of gradient update. Secondly, the exploratory process of candidate and evaluation is carried out by summing the momentum gradients of the two views and ranking them in descending order of saliency. In this process, selecting adversarial samples with stronger perturbation effects effectively improves the success rate of adversarial attacks. Finally, extensive experiments were conducted on three different datasets, and our generated adversarial samples were evaluated against contrastive learning models across two downstream tasks. The results demonstrate that the attack strategy proposed outperforms existing methods, significantly improving convergence speed. In the link prediction task, targeting the Cora dataset with perturbation rates of 0.05 and 0.1, the attack performance outperforms all baseline tasks, including the supervised baseline methods. The attack method is also transferred to other graph representation models, validating the method's strong transferability.


Assuntos
Algoritmos , Humanos , Aprendizado de Máquina
2.
Ying Yong Sheng Tai Xue Bao ; 35(1): 141-152, 2024 Jan.
Artigo em Chinês | MEDLINE | ID: mdl-38511450

RESUMO

Mixing native broadleaved tree species is a widely used method for renovating Pinus massoniana plantations. Soil microbial necromass carbon and organic carbon fractions are important parameters for evaluating the impacts of tree species mixing and soil organic carbon (SOC) stability. However, their responses to the mixing and renovation of P. massoniana plantation has not been understood yet. Here, we selected a pure P. massoniana plantation (PP) and a mixed P. massoniana and Castanopsis hystrix plantation, with ages of 16 (MP16) and 38 years (MP38), respectively, as the research objects. We quantified soil physical and chemical properties, microbial necromass carbon content, and organic carbon components at different soil layers to reveal whether and how the introduction of C. hystrix into P. massoniana plantation affected soil microbial necromass carbon and organic carbon components. The results showed that the mixed P. massoniana and C. hystrix plantation significantly reduced fungal necromass carbon content and the ratio of fungal/bacterial necromass carbon in the 0-20 cm and 20-40 cm soil layers. There were no significant differences in microbial necromass carbon contents, bacterial necromass carbon contents, and their contributions to SOC among the different plantations. The contribution of fungal necromass carbon to SOC was higher than that of bacterial necromass carbon in all plantation types. The contribution of soil mineral-associated organic carbon (MAOC) to SOC was higher than that of occluded particulate organic carbon (oPOC) and light-free particulate organic carbon (fPOC) for all plantation types. Mixing the precious broadleaved tree species (i.e., C. hystrix) with coniferous species (P. massoniana) significantly increased MAOC content and the contribution of MAOC, oPOC, and fPOC to SOC in the 0-20 cm and 20-40 cm soil layers. The MAOC of MP38 was significantly higher than that of PP in all soil layers and the MAOC of MP38 stands were significantly higher than MP16 stands in the 20-40 cm, 40-60 cm, and 60-100 cm soil layers, indicating that hybridization enhanced SOC stability and that the SOC of MP38 stands were more stable than MP16 stands. SOC and total nitrogen contents were the main environmental factors driving the changes in soil microbial necromass carbon, while soil total nitrogen and organically complexed Fe-Al oxides were the primary factors affecting organic carbon fraction. Therefore, SOC stability can be enhanced by introducing native broadleaved species, such as C. hystrix, during the management of the P. massoniana plantation.


Assuntos
Pinus , Árvores , Carbono/análise , Solo/química , Microbiologia do Solo , Nitrogênio/análise , Bactérias , China , Florestas
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA
...