GeoDL is Go Metals proprietary deep learning application to enhance a range of geological and geophysical map sets used in mining exploration.
Artificial intelligence and machine learning is a broad reference to computers mimicking human decision-making capabilities. Deep Learning is part of the broader family of machine learning methods based on artificial neural networks.
GeoDL learns from using multiple layers of data merged together to see combinations and patterns not readily available to the human eye. The primary layers are derived from structural, geophysical, geochemical, geographical and satellite images.
A digital image is composed of a set of small elements, called pixels. The signal strength of each pixel is represented by a number. GeoDL extracts features from the collection of numbers in a deep hierarchy. At the base level every feature represents characteristics of neighboring groups of pixels.
To capture less-localized, more-sophisticated information about the image, the algorithm repeats itself and creates depth. The feature map output from the first layer can be used as input to the second layer. The process then repeats.
These models are deep in the sense that they have many repeated layers. After extracting deep feature sets, GeoDL makes a final classification and outputs an enhanced map.
Better targets lead to less disturbance and improved ESG management.
Heat maps let companies focus on areas of interest and reduce exploration costs.