Skip to main content
All CollectionsVRIFY AIVRIFY AI Prediction Maps
Understanding: Set Up Learning Data
Understanding: Set Up Learning Data

Learn the key factors involved in setting: Set Up Learning Data in a VRIFY AI experiment.

Updated over a month ago

Overview

This section is where the targeted elements are defined, and where the learning data is separated into positive (mineralized) and negative (unmineralized) examples. The system uses learning points during training to recognize patterns in the data and make predictions. Each learning point helps improve the system's understanding of the underlying information.

Read on for more context and explanations about what this step entails.

For steps on configuring Set Up Learning Data, see here.

Key Concepts by Parameter

Parameter: Learning Points

  • The files in this drop-down are all available learning points files to choose from for your asset. If this drop-down is not available, it means you have only 1 file and it is being used by default.

Elevation Field

  • The values in this drop-down menu are populated from the column headers in your Learning Points shapefile attribute file. The target elevation selected here must match the column header that contains the Z coordinates of your learning points. This will be used to predict the 3D component of the prospectivity score.

Learning Data Filters

Sub-Parameter: Target Element(s)

  • The values in this drop-down menu are determined by your Learning Points shapefile attribute file.

  • You can set multiple Target Elements and Thresholds within one Prediction Map.

Sub-Parameter: Greater Than / Less Than Toggle

  • This toggle allows you to specify whether the target element must be greater than or less than the specified threshold in order to be considered a positive learning point.

  • By default, the filter is set to include elements greater than the corresponding target threshold. If necessary for your target, you can toggle the switch to change it to less than.

Sub-Parameter: Threshold

  • The threshold text box is where you indicate the grade the element must meet to be represented as a positive learning data point.

  • Element(s) that do not reach the defined grade threshold will be represented as a negative learning point.

  • The minimum and maximum values that can be used as a threshold will be displayed immediately under the Target Element / Threshold inputs. These are based on the lowest and highest values in your Learning Points shapefile attribute file.

Sub-Parameter: And/Or Toggle

  • When targeting multiple commodities, you can specify whether your targets are combined using “AND” logic or handled independently with “OR” logic.

  • “AND” Logic: All thresholds must be met in order to be considered a positive learning point.

    • Combines all targeted commodities into a single evaluation.

    • Positive and negative learning points are calculated collectively across all commodities.

    • This setting is useful when, for example, targeting a specific commodity while ensuring the absence of certain other elements.

  • “OR” Logic: Any of the target thresholds can be met in order to be considered a positive learning point.

    • Treats each targeted commodity independently, allowing separate evaluations for each with their own positive and negative learning points.

Learning Data Breakdown

  • This data table provides the ratio of positive and negative learning points based on the target thresholds configured in the Learning Data Filters.

  • This setting helps provide quantifiable context into how balanced your dataset is, and can be used to determine whether you may need to adjust the Negative Balance configuration in the Build Predictive Models step's advanced settings.

Still have questions?

Reach out to your dedicated VRIFY AI Contact or email Support@VRIFY.com for more information.

Did this answer your question?