机器学习在化学研究中的应用进展

Advance Application of Machine Learning in Chemical Research

机器学习算法的雏形在20世纪90年代就已出现,但是局限于当时的硬件条件,计算机寻找关联数据间复杂函数的能力有限,机器学习算法在当时并没有得到长足的发展。随着计算机科学的飞速发展,电脑软件与硬件性能得到了巨大的提升,前人设想的基于机器学习的人工智能算法一定程度上得到了实现。现阶段,机器学习算法已经可以从大批量的观测数据中广泛的寻找变量关系,进而帮助人们发现更多的特定规律。作为一种新兴的方法,在化学领域中,机器学习算法可以高效利用并处理高通量实验产生的大批量数据。近年来,各种算法工具不断完善,使用机器学习算法的门槛不断降低,越来越多的研究员将机器学习应用于化学研究,开启了化学研究的新篇章。本文通过介绍了机器学习的发展过程,机器学习应用的简单方法,机器学习应用化学研究三个方面,简单总结了现阶段机器学习在化学研究中的应用进展。

The prototype of machine learning algorithm has appeared since the 1990s,but it was limited by the hardware conditions at that time,so the ability of computers to find complex functions between related data was limited,and the machine learning algorithm did not get a great development at that time.With the rapid development of computer science,the performance of computer software and hardware has been greatly improved,and the artificial intelligence algorithm based on machine learning envisioned by predecessors has been realized to a certain extent.At present,machine learning algorithms have been able to extensively search for variable relationships from large quantities of observed data to help people discover more specific laws.As an emerging method,machine learning algorithms can efficiently utilize and process large quantities of data generated by high-throughput experiments in the field of chemistry.In recent years,a variety of algorithm tools are constantly improved,and the threshold of using machine learning algorithm is constantly lowered.More and more researchers apply machine learning to chemical research,opening a new chapter in chemical research.By introducing the development process of machine learning,the application of machine learning simple methods,machine learning applied chemistry research,the current stage of machine learning in the application of chemical research was summarized.