RQ2: Can we train different mutation models?

We report here data and results for Cluster-Models:

  • Datasets & Predictions

  • Clustering & Silhouette values

Datasets & Predictions

We share the datasets and predictions of the models. Each dataset is partitioned in 80% training, 10% validation, and 10% testing set. The predictions of the models represent the mutant generated for the 10% test set.

Clustering - Silhouette

Download data - script to compute silhouette values as well as the results of the distribution (boxplots).