Figure 1.

Example of ML pipeline for TGx data. Data Acquisition and Preprocessing: Data is collected and analyzed to ensure the quality of the dataset. During the preprocessing, feature selection and/or feature transformations may be applied to improve stability. Training-Hyperparameter tuning-Validation loop: candidate models are fit to the data. This is embedded in an iterative process where for each candidate model the best hyperparameters are optimized through the validation step. Model Selection and Testing: Optimized candidate models are identified and the best ones are tested on a final hold-out dataset to evaluate generalization capabilities.