2317 deposits showed lower probability because they were subordinated to the main deposit type that the algorithm was designed to model. This underscores the importance of careful algorithm design to capture the specific characteristics of each district. The methodology helped overcome exploration challenges related to dense vegetation by improving the correlation between predictions and field observations. In collaboration with IAMGOLD Perú, priority zones within the study area were identified and validated through sampling and spectral analysis, although detailed results remain confidential. The same approach was also applied to a lithium exploration project targeting pegmatites in the Pontiac Group in Abitibi, Canada, where radiometric geophysics successfully mapped targets beneath dense forest cover and showed excellent correlation with field data, demonstrating the transferability of the method. 5. DISCUSSION The results highlight the strengths and limitations of AI-driven prospectivity modeling for mineral exploration. High-prospectivity zones that coincide with known deposits build confidence in the methodology and support its use as a decision-support tool. False positives can be interpreted as areas where further data acquisition or ground-truthing is required, rather than outright errors, and therefore help prioritize field work and data collection. Key enablers of the workflow include access to diverse geoscientific datasets, robust data preprocessing, and close collaboration with industry partners such as IAMGOLD Perú. Barriers include data confidentiality, uneven data coverage, and computational costs associated with large geospatial models. Despite these limitations, the method is transferable to other regions with similar deposit styles, provided that appropriate local calibration is undertaken. By integrating multiple sources of information and explicitly quantifying uncertainty, the workflow supports more transparent and defensible exploration decisions. This contributes to building trust with stakeholders, accelerating the identification of high-potential areas, and enabling more efficient use of exploration budgets, aligned with the themes of trust, transformation, and technology promoted by WMC 2026. 6. CONCLUSIONS AND IMPLICATIONS FOR INDUSTRY The study confirms that Fuzzy Logic and AI-based techniques are effective tools for mineral exploration because they integrate complex datasets and reduce traditional interpretation biases. The multifactorial approach mitigated environmental challenges such as dense vegetation and rugged topography, producing predictions that are more accurate and relevant for strategic decision-making. By generating high-probability targets through AI and Machine Learning, the workflow optimizes exploration planning and resource allocation, reducing risk and cost. For industry, this translates into shorter timelines from regional screening to drill targeting, more focused investment decisions, and improved ability to deliver the minerals society urgently needs in a responsible manner. The application of advanced methodologies demonstrates a continuous commitment to innovation and the improvement of exploration practices. Although this study focused on Fuzzy Logic, future work may compare additional AI techniques to broaden the scope and effectiveness of prospectivity models. Despite limitations related to data confidentiality and computational cost, the methodology has proven to be versatile and scalable, with potential applications in diverse geological contexts and commodities. ACKNOWLEDGEMENTS Special thanks are extended to IAMGOLD Perú S.A.C. for their support in validating the results through field work.
RkJQdWJsaXNoZXIy MTM0Mzk2