*5.1. Machine Learning Models Development*

The machine learning models were trained to predict Estatic as a function of the RHOB, DTs, and DTc. The training data were collected from Well-A. Figure 5 compares the actual and estimated Estatic for the training dataset.

**Figure 5.** Actual and estimated static Young's modulus (Estatic) for the training dataset collected from Well-A.

Figure 5 shows that all machine learning models predicted Estatic with very high accuracy. M-FIS predicted Estatic with AAPE of only 0.05% and R of 0.999995. FNN model estimated the Estatic with AAPE and R of 0.78% and 0.999491, respectively, while SVM model estimated Estatic with AAPE of 0.55% and R of 0.999634, and the ANN model predicted Estatic with AAPE of 0.98% and R of 1.000000. The good matching between the actual and estimated Estatic for the training dataset shown in Figure 5 proves the high accuracy of the machine learning models in evaluating Estatic.
