排序方式： 共有2条查询结果，搜索用时 0 毫秒
1
1.
近年来,随着AI for science(科学智能)的蓬勃发展,人工智能与各门学科的交叉融合逐渐成为一个显著的科学研究趋势。但是,AI for science所涉及的范围很广、学科众多,因此,将其梳理成一个统一的体系能够更好地为初入领域的研究者导航。本文认为,尽管每一门学科研究的对象、方法看似千差万别,但人工智能可以为科学研究提供一套普适的范式和方法,解决各科学领域内的重要问题。本文将从科学仿真、设计和控制、发现三个方面展开综述,明确任务设置,梳理当前的代表性工作,并通过具体的例子,阐释人工智能如何为科学研究助力,以期使读者能够更好应用已有的方法,或者研究新的方法。 相似文献
2.
Recent extensive studies of Escherichia coli (E. coli) chemotaxis have achieved a deep understanding of its mi- croscopic control dynamics. As a result, various quantitatively predictive models have been developed to describe the chemotactic behavior of E. coli motion. However, a population-level partial differential equation (PDE) that rationally incorporates such microscopic dynamics is still insufficient. Apart from the traditional Keller-Segel (K-S) equation, many existing population-level models developed from the microscopic dynamics are integro-PDEs. The difficulty comes mainly from cell tumbles which yield a velocity jumping process. Here, we propose a Langevin approximation method that avoids such a difficulty without appreciable loss of precision. The resulting model not only quantitatively repro- duces the results of pathway-based single-cell simulators, but also provides new inside information on the mechanism of E. coli chemotaxis. Our study demonstrates a possible alternative in establishing a simple population-level model that allows for the complex microscopic mechanisms in bacterial chemotaxis. 相似文献
1