2. Data framing and preparation

2.1 Principle

The scoping and data preparation phase is an essential prerequisite for building a high-performance ML model. In a branch of engineering as specific as geotechnics, it is crucial to clearly define the scope of the project and meticulously prepare the data before starting model training. The aim of this step is to structure data from multiple sources, resolve inconsistencies and ensure that it matches algorithm specifications.

The scoping process includes identifying key variables and designing a data architecture that will facilitate their exploitation. At the same time, careful preparation of the data is necessary to ensure that it is clean and ready to be used efficiently by the ML models. By carrying out these steps correctly, we maximize the...