{"title":"Machine Learning in Petrology: State-of-the-Art and Future Perspectives","authors":"Maurizio Petrelli","doi":"10.1093/petrology/egae036","DOIUrl":null,"url":null,"abstract":"The present manuscript reports on the state-of-the-art and future perspectives of Machine Learning (ML) in petrology. To achieve this goal, it first introduces the basics of ML, including definitions, core concepts, and applications. Then, it starts reviewing the state-of-the-art of ML in petrology. Established applications mainly concern the so-called data-driven discovery and involve specific tasks like clustering, dimensionality reduction, classification, and regression. Among them, clustering and dimensionality reduction have been demonstrated to be valuable for decoding the chemical record stored in igneous and metamorphic phases and to enhance data visualization, respectively. Classification and regression tasks find applications, for example, in petrotectonic discrimination and geo-thermobarometry, respectively. The main core of the manuscript consists of depicting emerging trends and the future directions of ML in petrological investigations. I propose a future scenario where ML methods will progressively integrate and support established petrological methods in automating time-consuming and repetitive tasks, improving current models, and boosting discovery. In this framework, promising applications include (a) the acquisition of new multimodal petrologic data, (b) the development of data fusion techniques, physics-informed ML models, and ML-supported numerical simulations, and (c) the continuous exploration of the ML potential in petrology. To boost the contribution of ML in petrology, our main challenges are: (a) to improve the ability of ML models to capture the complexity of petrologic processes, (b) progressively link machine learning algorithms with the physical and thermodynamic nature of the investigated problems, (c) to start a collaborative effort among researchers coming from different disciplines, both in research and teaching.","PeriodicalId":16751,"journal":{"name":"Journal of Petrology","volume":"117 1","pages":""},"PeriodicalIF":3.5000,"publicationDate":"2024-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Petrology","FirstCategoryId":"89","ListUrlMain":"https://doi.org/10.1093/petrology/egae036","RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GEOCHEMISTRY & GEOPHYSICS","Score":null,"Total":0}
引用次数: 0
Abstract
The present manuscript reports on the state-of-the-art and future perspectives of Machine Learning (ML) in petrology. To achieve this goal, it first introduces the basics of ML, including definitions, core concepts, and applications. Then, it starts reviewing the state-of-the-art of ML in petrology. Established applications mainly concern the so-called data-driven discovery and involve specific tasks like clustering, dimensionality reduction, classification, and regression. Among them, clustering and dimensionality reduction have been demonstrated to be valuable for decoding the chemical record stored in igneous and metamorphic phases and to enhance data visualization, respectively. Classification and regression tasks find applications, for example, in petrotectonic discrimination and geo-thermobarometry, respectively. The main core of the manuscript consists of depicting emerging trends and the future directions of ML in petrological investigations. I propose a future scenario where ML methods will progressively integrate and support established petrological methods in automating time-consuming and repetitive tasks, improving current models, and boosting discovery. In this framework, promising applications include (a) the acquisition of new multimodal petrologic data, (b) the development of data fusion techniques, physics-informed ML models, and ML-supported numerical simulations, and (c) the continuous exploration of the ML potential in petrology. To boost the contribution of ML in petrology, our main challenges are: (a) to improve the ability of ML models to capture the complexity of petrologic processes, (b) progressively link machine learning algorithms with the physical and thermodynamic nature of the investigated problems, (c) to start a collaborative effort among researchers coming from different disciplines, both in research and teaching.
本手稿报告了岩石学中机器学习(ML)的最新进展和未来展望。为实现这一目标,本文首先介绍了机器学习的基础知识,包括定义、核心概念和应用。然后,开始回顾机器学习在岩石学中的最新应用。已有的应用主要涉及所谓的数据驱动发现,并涉及聚类、降维、分类和回归等具体任务。其中,聚类和降维已被证明对解码火成岩相和变质岩相中存储的化学记录以及增强数据可视化分别具有重要价值。分类和回归任务分别应用于岩石构造判别和地质测温等领域。手稿的主要核心内容包括描绘 ML 在岩石学研究中的新兴趋势和未来方向。我提出的未来设想是,ML 方法将逐步整合并支持既有的岩石学方法,实现耗时和重复性任务的自动化,改进现有模型并促进发现。在此框架下,有前景的应用包括:(a) 获取新的多模态岩石学数据;(b) 开发数据融合技术、物理信息 ML 模型和 ML 支持的数值模拟;(c) 不断探索 ML 在岩石学中的潜力。为了提高 ML 在岩石学中的贡献,我们面临的主要挑战是(a) 提高 ML 模型捕捉岩石学过程复杂性的能力,(b) 逐步将机器学习算法与所研究问题的物理和热力学性质联系起来,(c) 启动来自不同学科的研究人员在研究和教学方面的合作努力。
期刊介绍:
The Journal of Petrology provides an international forum for the publication of high quality research in the broad field of igneous and metamorphic petrology and petrogenesis. Papers published cover a vast range of topics in areas such as major element, trace element and isotope geochemistry and geochronology applied to petrogenesis; experimental petrology; processes of magma generation, differentiation and emplacement; quantitative studies of rock-forming minerals and their paragenesis; regional studies of igneous and meta morphic rocks which contribute to the solution of fundamental petrological problems; theoretical modelling of petrogenetic processes.