Table 2.

Preprocessing Pipeline – algorithms, libraries, and parameter settings.

Preprocessing stage	Method / Algorithm	Library / Tool	Parameter Settings / Values
Missing value handling	Median imputation (Numerical)	numpy.median()	Replaces missing values with column-wise median
Normalization	Min–max normalization	sklearn.preprocessing.MinMaxScaler	Feature range: [0, 1]
Noise removal	Savitzky-golay filter	scipy.signal.savgol_filter()	window_length = 11, polyorder = 3, mode = ‘interp’
Data augmentation	SMOTE (Synthetic Minority Oversampling Technique)	imblearn.over_sampling.SMOTE	k_neighbors = 5, sampling_strategy = ‘auto’, random_state = 42