Normalizes data using MinMax or Standard scaling. Args: X (np.ndarray): Input data. method (str): 'minmax' or 'standard'.

Saves metrics to a file in JSON or CSV format. Args: metrics_dict (dict): Dictionary of metrics. filepath (str): Output file

Plots the distribution of labels in a pie chart.

Builds and compiles a Keras Sequential model. Args: input_shape (int): Number of input features. architecture (str): One of

Prints classification report.

Converts pandas DataFrames or Series to numpy arrays.

Converts scaled data back to original scale.

Loads a trained Keras model.

Loads raw transaction data from a CSV file.

Loads input data for prediction. Args: from_csv (bool): If True, loads from a CSV file. csv_path (str): Path to CSV file (mu

Logs training or evaluation metrics in formatted output. Args: metrics_dict (dict): Dictionary of metric name to value. pref

Normalizes numeric columns from CSV and writes output.

Plots a confusion matrix heatmap.

Plots training and validation accuracy/loss curves.

Selects features, encodes categoricals, scales numerics.

Runs model inference on input data. Args: model: Trained Keras model. data (np.ndarray): Input features. threshold (

Saves the scaler to a file.

Splits data for simulated federated learning clients.

Splits data into train and test sets.

Attempts a stratified split to preserve label proportions across clients.

Trains the model on local data. Args: model: Compiled Keras model. X (np.ndarray): Features. y (np.ndarray): Labels.

Validates that the required columns exist in the dataset.

Evaluates a model and prints metrics. Args: model: Trained model (Keras or scikit-learn). X_test (array-like): Test features

Loads a scaler from file.

Splits dataset into federated clients. Args: X (array-like): Features. y (array-like): Labels. num_clients (int): Nu