PRACTICE

Learning thru Observations

Objectives & Learning Data

Learning objectives may include any target of interest or unwanted event such as asset incidents, failures or interruptions. Domain experts identify potentially influencing data which is then integrated using geospatial and linear data management tools in preparation for the learning process.

Methods

The learning process determines the "best" Inferential statistical, supervised or unsupervised method for the given use case. More than a hundred methods are available and the learning process determines the method most appropriate for the given criteria, transparency and required level of performance.

Validation

The "models" resulting from the learning methods process are validated thru cross-validation and testing with unseen observations. Prediction results are then uniquely associated with levels of confidence and other performance measures.

Model Application

Models meeting criteria of performance and transparency are then applied to assets of similar types resulting in validated prediction results.

Results & Actions

Results are normally expressed as levels of susceptibility or probability of the target of interest. These results are then monetized to support mitigation decision-making and maintenance and capital planning.