141 PREDICTION OF PORPHYRY-TYPE DEPOSIT OCCURRENCE IN YUKON (CANADA) USING CONVOLUTIONAL VARIATIONAL AUTOENCODERS (VAE-CNN) *N.H. Berrospi1 1Department of Geological Engineering, National University of Engineering, Peru, (*Presenting author: neilberrospir@uni.pe) ABSTRACT The increasing global demand for critical metals driven by the energy transition necessitates more efficient mineral exploration strategies, particularly in remote and underexplored regions. Traditional exploration methods are costly, time-intensive, and often constrained by accessibility. While supervised machine learning techniques have been introduced to enhance targeting efficiency, they rely heavily on labeled datasets, which may introduce bias and false negatives in poorly explored terrains. This study presents an unsupervised deep learning approach based on Convolutional Variational Autoencoders (VAE-CNN) to predict porphyry-type deposit occurrence in the Yukon. The model is trained on multi-source geoscientific raster data without requiring explicit non-deposit labels. Instead, it learns latent representations capturing geological signatures associated with porphyry systems. A total of six raster layers, including geophysical, topographic, and remote sensing data, were integrated at a spatial resolution of 100 m. Subsurface patterns were extracted using a sliding window and data augmentation strategy, generating approximately 9,600 training samples. The learned latent space was used to derive a representative signature from subcrops associated with the 25 known deposits in the western sector, which was then compared across the study area using Euclidean distance to generate a prospectivity map. Validation using 60 withheld deposits from the eastern sector resulted in 49 correctly identified targets above a similarity threshold of 0.80, demonstrating the model’s effectiveness. In addition, the prospectivity map delineated 30 new high-prospectivity zones (similarity ≥ 0.90), representing potential exploration targets with strong resemblance to the learned porphyry signature. The proposed methodology provides a data-driven framework for mineral prospectivity mapping, reducing reliance on labeled data and minimizing the risk of false negatives. KEYWORDS Unsupervised Learning, Variational Autoencoder, Mineral Exploration, Porphyry Deposits, Yukon, Prospectivity Mapping 1. INTRODUCTION The global transition toward low-carbon energy systems has significantly increased the demand for critical metals, particularly copper, which plays a fundamental role in electrification, renewable energy infrastructure, and energy storage technologies. According to the International Energy Agency, demand for copper and other energy transition minerals is expected to grow substantially in the coming decades, placing increasing pressure on the mineral exploration sector to identify new economically viable deposits (IEA, 2021). Despite this growing demand, mineral exploration remains a high-risk and capital-intensive process, especially in remote and underexplored regions such as the Yukon. Traditional exploration methods, including geological mapping, geochemical sampling, and geophysical surveys, have historically been effective in discovering mineral systems. However, these approaches are often constrained by high operational costs, logistical challenges, and limited spatial coverage, particularly in areas with difficult access or sparse infrastructure (Singer & Kouda, 1999).
RkJQdWJsaXNoZXIy MTM0Mzk2