Scholarlink

One of the principal areas of physical chemistry is the investigation of the structure–property relationshipQuantitative structure-activity/property relationship. This encompasses the correlation of a property with the characteristics of interest to a specific field of chemistry and materials science. Structure–property modeling strives to create data-driven models capable of forecasting the properties of chemical compounds based on their structure. These predictive models serve as valuable aids to experimentalists and can aid in the discovery of substances and the development of materials with predetermined properties. Specialized commercial applications tailored for such predictions are often costly and frequently beyond the reach of academic research laboratories. Nevertheless, in recent years, freely accessible web services have been increasingly developed to forecast the physicochemical characteristics of substances and materials. However, recognising the vastness of all possible options, this chapter does not cover all existing services but focuses solely on those that, in the subjective view of the authors, fulfill the criteria of accuracy, public accessibility, and user-orientation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning-Driven Web Tools for Predicting Properties of Materials and Molecules

  • 章节

  • 发表日期：

    2025年

查看全文
  • Dmitry M. Makarov,
  • Pavel S. Bocharov,
  • Michail M. Lukanov,
  • Alexander A. Ksenofontov

摘要

One of the principal areas of physical chemistry is the investigation of the structure–property relationshipQuantitative structure-activity/property relationship. This encompasses the correlation of a property with the characteristics of interest to a specific field of chemistry and materials science. Structure–property modeling strives to create data-driven models capable of forecasting the properties of chemical compounds based on their structure. These predictive models serve as valuable aids to experimentalists and can aid in the discovery of substances and the development of materials with predetermined properties. Specialized commercial applications tailored for such predictions are often costly and frequently beyond the reach of academic research laboratories. Nevertheless, in recent years, freely accessible web services have been increasingly developed to forecast the physicochemical characteristics of substances and materials. However, recognising the vastness of all possible options, this chapter does not cover all existing services but focuses solely on those that, in the subjective view of the authors, fulfill the criteria of accuracy, public accessibility, and user-orientation.