Abstract:In various geosciences branches, including mineral exploration, geometallurgical characterization on established mining operations, and remote sensing, the regionalized input variables are spatially well-sampled across the domain of interest, limiting the scope of spatial uncertainty quantification procedures. In turn, response outcomes such as the mineral potential in a given region, mining throughput, metallurgical recovery, or in-situ estimations from remote satellite imagery, are usually modeled from a much-restricted subset of testing samples, collected at certain locations due to accessibility restrictions and the high acquisition costs. Our limited understanding of these functions, in terms of the multi-dimensional complexity of causalities and unnoticed dependencies on inaccessible inputs, may lead to observing changes in such functions based on their geographical location. Pooling together different response functions across the domain is critical to correctly predict outcome responses, the uncertainty associated with these inferred values, and the significance of inputs in such predictions at unexplored areas. This paper introduces the notion of a dual random field (dRF), where the response function itself is considered a regionalized variable. In this way, different established response models across the geographic domain can be considered as observations of a dRF realization, enabling the spatial inference and uncertainty assessment of both response models and their predictions. We explain how dRFs inherit all the properties from classical random fields, allowing the use of standard Gaussian simulation procedures to simulate them. These models are combined to obtain a mineral potential response, providing an example of how to rigorously integrate machine learning approaches with geostatistics.

What problem does this paper attempt to address?

This paper attempts to solve the problem that in Mineral Potential Mapping (MPM), due to the complexity of geographical areas and the non - uniformity of data distribution, traditional methods are difficult to accurately predict and evaluate the mineral potential of unexplored areas. Specifically, the paper mainly focuses on the following aspects: 1. **Limitations of traditional methods**: Existing statistical and machine - learning methods, when dealing with MPM problems, often overlook the spatial dependence relationships among input variables and the spatial relationships between input and output. This may lead to over - estimation of known mineralized areas and under - estimation of unknown resources. 2. **Challenges of spatial variability**: Causal factors in different regions may be significantly different, and using a single global model may be too rigid to adapt to these changes. Therefore, a method that can adapt to different regional characteristics is needed to improve the accuracy of prediction. 3. **Need for multi - model fusion**: In order to better capture the changing characteristics of different regions, the paper proposes a method of combining multiple local response models to achieve more accurate spatial inference and uncertainty assessment. For this purpose, the paper introduces the concept of **dual random field (dRF)**, in which the response function itself is regarded as a regionalized variable. In this way, response models at different geographical locations can be regarded as different realizations of dRF, thus allowing for spatial inference and uncertainty assessment of response models and their predictions. ### Main contributions - **dRF framework**: The paper proposes a mathematical framework that enables different response models to perform global and local statistical inferences within geographical regions. - **Simplification of linear response functions**: By restricting to linear response functions, the research is simplified to study random fields defined on unit hyperspheres, which helps to keep the usual geostatistical concepts such as mathematical expectation and covariance controllable. - **Application of SVM random fields**: It shows how to use Support Vector Machine (SVM) to build local linear models and combine these models into a unified dRF. - **Case study**: Through a practical case study, it shows how to apply the dRF method for MPM of nickel deposits, using different geophysical input features and nickel geochemical data obtained from shallow layers. In conclusion, this paper aims to provide a new method to deal with the problems of spatial uncertainty and complexity in MPM by introducing the dRF framework, thereby improving the prediction accuracy of the mineral potential in unexplored areas.

Dual Random Fields and their Application to Mineral Potential Mapping

3D Mineral Prospectivity Mapping with Random Forests: A Case Study of Tongling, Anhui, China

A Truly Spatial Random Forests Algorithm for Geoscience Data Analysis and Modelling

Mineral Prospectivity Mapping based on Isolation Forest and Random Forest: Implication for the Existence of Spatial Signature of Mineralization in Outliers

An Uncertainty-Quantification Machine Learning Framework for Data-Driven Three-Dimensional Mineral Prospectivity Mapping

Revealing Geochemical Patterns Associated with Mineralization Using t-Distributed Stochastic Neighbor Embedding and Random Forest

Lithologic mapping using Random Forests applied to geophysical and remote-sensing data: A demonstration study from the Eastern Goldfields of Australia

Random forest as a generic framework for predictive modeling of spatial and spatio-temporal variables

Quantifying uncertainty and improving prospectivity mapping in mineral belts using transfer learning and Random Forest: A case study of copper mineralization in the Superior Craton Province, Quebec, Canada

A Segmentation Approach for Stochastic Geological Modeling Using Hidden Markov Random Fields

Modulating the Impacts of Stochastic Uncertainties Linked to Deposit Locations in Data-Driven Predictive Mapping of Mineral Prospectivity

Exploring substitution random functions composed of stationary multi-Gaussian processes

GeoRF: a geospatial random forest

Multivariate statistical analysis and bespoke deviation network modeling for geochemical anomaly detection of rare earth elements

A multivariate Gaussian random field prior against spatial confounding

Binary Spatial Random Field Reconstruction from Non-Gaussian Inhomogeneous Time-series Observations

Conditional formulation for the calibration of multi-level random fields with incomplete data

Random Spatial Forests

Dual Neural Network Architecture for Determining Epistemic and Aleatoric Uncertainties

Uncertainty Quantification by Random Measures and Fields

REDS: Random Ensemble Deep Spatial prediction