摘要
由一名新闻记者-机器人与机器学习每日新闻的工作人员新闻编辑-研究人员详细介绍了机器学习的新数据。根据NewsRx记者从德国明斯特发回的新闻报道，Church说：“将分子结构编码成计算机可读、可利用的格式是所有化学科学中机器学习应用的关键步骤。根据应用的不同，当前的表示形式在复杂性和形状上有很大差异。”这项研究的财政支持来自德国研究基金会(DFG)。新闻记者从蒙斯特大学的研究中得到一句话：“因此，领域特定表征的数量正在迅速增长，有些表征不断被改变和重复，这些定制的表征增加了进入和方法适应的障碍，从而减缓了应用的进展。”我们提出了一种通用算法，能够仅基于给定的数据集产生高度特异性的表示。该算法利用结构查询和进化方法生成可插入的分子指纹。这些指纹非常适合分子机器筛选，能够准确预测反应性、性质和生物活性。我们展示了其固有的可解释性，允许提取知识，如反应趋势。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News – Researchers detail new data in Machine Learning. According to news reporting from Munster, Germany, by NewsRx journalists, resear ch stated, “Encoding molecular structures into a computer - readable, utilizable format is the key step for any machine learning application in all chemical scie nces. Current representations vary strongly in complexity and shape, depending o n the application.” Financial support for this research came from German Research Foundation (DFG). The news correspondents obtained a quote from the research from the University o f Munster, “Therefore, the number of domain -specific representations is rapidly growing, with some being altered and retuned constantly. These tailored represe ntations raise the barriers for entry and method adaption, thus decelerating pro gress in application. Herein, we present a general algorithm capable of yielding a highly specific representation solely based on a given dataset. The algorithm utilizes structural queries and evolutionary methodologies to generate interpre table molecular fingerprints. These are highly suited for molecular machine lear ning, enabling the accurate prediction of reactivity, property, and biological a ctivity. We demonstrate its native interpretability, allowing for the extraction of knowledge, such as reactivity trends.”
NETL
NSTL