Table 2.

ML Software Setup [Abbreviations: ML = Machine Learning; GPU = Graphics Processing Unit; HPC = High-Performance Computing; DFFNN = Deep Feed Forward Neural Network; SD = Standard Deviation; LR = Learning Rate].

Aspect	Description
Technical environment	PC with sufficient hardware; no grid of GPU or HPC system
Data format requirements	Table based, e.g., CSV format
Data import	Use case specific; manual import; standard normalization and error handling
ML technology	TensorFlow [24]; no hyperparameter optimization framework
Model	DFFNN [29] of different shapes as regressor
Quality metrics	$MS{E}_{\mathrm{arr}}$ based on [19], see Formula (1); SD of $MS{E}_{\mathrm{arr}}$
Sanity check	$RS{W}_{\mathrm{input}}$, see Formula (2)
Result optimization objective	Minimize $MSE$, $MS{E}_{\mathrm{arr}}$ respectively
Optimization space	$LR$, $f_{\mathrm{ModelSpread}}$, $n_{\mathrm{ModelDepth}}$, $t_{\mathrm{alg}}$