Investigating hardware and software aspects in the energy consumption of machine learning: A green AI-centric analysis

CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE(2023)

引用 0|浏览2
暂无评分
摘要
Management of storage pest insects relies heavily on chemical control, and there is need to develop more sustainable management practices. Here, we evaluated the impact of 2% ethanol plant extracts of Ajuga reptans L., Ajuga pyramidalis L. (both Lamiaceae), Urtica dioica L. (Urticaceae), and Cannabis sativa L. (Cannabaceae) plants mortality, population growth, and developmental stability [measuring fluctuating asymmetry (FA)] of the rice weevil, Sitophilus oryzae (L.) (Coleoptera: Curculionidae), a worldwide stored product pest. FA refers to small, random deviations occurring be-tween the left and right sides of bilaterally symmetrical organisms; these deviations increase in response to environmental stress, making FA a reliable method to meas-ure the impact of stress. FA was measured by means of geometric morphometrics, method that allows for analyzing the whole landmark configuration of the insect, rather than taking single measurements. Extracts of the mentioned plants were used treat maize (Zea mays L., Poaceae) kernels on which experimental populations the rice weevil were grown, and we assessed mortality after 24- 72 h, population growth after 30- 90 days, and developmental stability after 90 days. Screening bioas-says showed that S. oryzae adults were most affected by Ajuga extracts; Ajuga spp., especiallyA. reptans, significantly reduced population growth. In concordance, Ajuga extracts increased FA. The effects of U. dioica and C. sativa extract were less pro-nounced. None of the extracts significantly affected insect mortality. Overall, it can be
更多
查看译文
关键词
algorithm,ARM,artificial intelligence,CO2,energy,green AI,machine learning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要