Description
GeoStru Artificial Intelligence – GILA allows local training of Machine Learning models for the estimation of geotechnical parameters. Machine Learning is a subset of Artificial Intelligence that enables models to learn from data.
The program creates a model for each parameter and attempts to learn the most commonly used correlations within penetration test data, which will be provided as input.
The software allows training of Machine Learning models directly from .dip or .dipx files, which can be generated by the Dynamic Probing software. Alternatively, models can be trained from a .csv file.
Furthermore, the program provides the following functionalities:
- Assigning a lithological classification for each soil type present in the input files.
- Performing parameter estimations using the trained models.
- Saving trained models locally for future reuse.
- Conducting quantitative and graphical analyses of the performance of the trained models.