Distributed training is the process of partitioning the workload across multiple processing units to speed up the training especially in ML. [Read More]

Image analysis extracts feature information to quantify shapes, enumerate object’s structures and characterize shape of structures. Rock particles result from erosion, blasting, comminution, and others. The geometric characteristics of these particles encode information about their generating process. The mineralogy influences mechanical properties and determines the shape and size of individual... [Read More]

Exploration projects collect information over the years. The information changes due to incoming data, change of personnel, among other factors. Changes in the data must be justified and well documented, it requires a sound database and professionals to provide insightful details. A simple case is to understand the impact of... [Read More]

What is a compute in Azure ML Compute instances in Azure ML Studio are single-owned cloud-based workstations used as a pre configured, optimized and fully customizable environment for ML applications. They also allow us to share files between instances and built manageable applications. [Read More]

Sometimes we deal with sub-blocked models generated with a software that is not available to us, and we have access only to GSLIB. When using GSLIB’s programs, we must consider: (1) they use structured unrotated grids, and (2) write to disk. This may not discourage to practitioners as it’s fairly... [Read More]