Figure 4 - An Artificial Neural Network (ANN) model to predict the parasitic capacitance based on the etch depth, tilt and source sigma. The prediction accuracy on the test data was found to be 99.8%. The metric to measure the difference between the predicted capacitance and the actual capacitance was absolute percent error (APE). 75% percent of test cases had an APE of 0.28% or lower. An accurate machine learning model allows a smaller parameter space to be explored and hence lower time and computational resources.