Scholarlink

<p>Uranium (U), a naturally occurring radioactive element in the Earth's crust, presents considerable hazards to the environment and human health. This study primarily aimed to probabilistically model uranium (U) concentrations based on three key influencing factors. In this study, a combination of statistical methods and advanced artificial intelligence techniques was employed to analyze the data and model the relationships between uranium and other geochemical elements. To achieve this, Monte Carlo Simulation (MCS) was employed to evaluate U under various conditions. Dysprosium, Erbium, Lanthanum, Europium, Lutetium, Neodymium, Praseodymium, Thulium, Yttrium, and Ytterbium were selected as input variables for the model. Initially, two predictive equations were developed using statistical and artificial intelligence methods, which were subsequently applied in the MCS framework to simulate U concentrations. The results demonstrated that these approaches are capable of accurately simulating U distributions. The mean value of U obtained through MCS was 240&#xa0;ppm, closely matching the measured mean of 237&#xa0;ppm and its concentration at a 90% confidence level does not exceed 487&#xa0;ppm.</p>

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

Feasibility of Monte Carlo simulation and comparison of statistical and artificial intelligence methods for predicting uranium in Baghak, NE Iran

  • 期刊论文

  • 发表日期：

    2026年3月28日

查看全文
  • Z. Varmazyari,
  • S. S. Ghannadpour,
  • H. Katibeh

摘要

Uranium (U), a naturally occurring radioactive element in the Earth's crust, presents considerable hazards to the environment and human health. This study primarily aimed to probabilistically model uranium (U) concentrations based on three key influencing factors. In this study, a combination of statistical methods and advanced artificial intelligence techniques was employed to analyze the data and model the relationships between uranium and other geochemical elements. To achieve this, Monte Carlo Simulation (MCS) was employed to evaluate U under various conditions. Dysprosium, Erbium, Lanthanum, Europium, Lutetium, Neodymium, Praseodymium, Thulium, Yttrium, and Ytterbium were selected as input variables for the model. Initially, two predictive equations were developed using statistical and artificial intelligence methods, which were subsequently applied in the MCS framework to simulate U concentrations. The results demonstrated that these approaches are capable of accurately simulating U distributions. The mean value of U obtained through MCS was 240 ppm, closely matching the measured mean of 237 ppm and its concentration at a 90% confidence level does not exceed 487 ppm.