Abstract Geochemical data are widely regarded as valuable for mineral exploration because they can be used to recognize ore-related geochemical anomalies. Many techniques have been used to identify geochemical anomalies within geochemical datasets, but most of these techniques focus on the concentration or fractal characteristics of one or a combination of indicator element(s). Indicator elements for a particular type of deposit are selected based on expert knowledge, but this approach is subjective and valuable information associated with the elements that are not selected is lost. This study describes a new method that maximize the extraction of valuable information from all elements available in a geochemical dataset. The technique involves geochemical metallogenic potential mapping based on cluster analysis (GMPM-CA), which enables the identification of prospective areas suitable for further mineral exploration. The results show that GMPM-CA can produce a map of mineralization potential with excellent predictive capacity. Furthermore, GMPM-CA is a practical method that does not require expert geological or geochemical knowledge, and can be integrated with existing geochemical anomaly identification methods.

中文翻译：

---

基于聚类分析的地球化学成矿潜力：一种从地球化学数据中提取矿产勘探有价值信息的新方法

摘要 地球化学数据可用于识别与矿石有关的地球化学异常，因此被广泛认为对矿产勘探具有重要价值。许多技术已被用于识别地球化学数据集中的地球化学异常，但大多数这些技术侧重于一种或多种指示元素的浓度或分形特征。特定类型矿床的指标元素是根据专家知识选择的，但这种方法是主观的，与未选择的元素相关的有价值的信息会丢失。这项研究描述了一种新方法，可以最大限度地从地球化学数据集中的所有可用元素中提取有价值的信息。该技术涉及基于聚类分析的地球化学成矿潜力图（GMPM-CA），这使得能够确定适合进一步矿产勘探的潜在区域。结果表明，GMPM-CA 可以生成具有出色预测能力的矿化潜力图。此外，GMPM-CA 是一种实用的方法，不需要专业的地质或地球化学知识，可以与现有的地球化学异常识别方法相结合。