plot_integrated_gradients {MLwrap} | R Documentation |
Plotting Integrated Gradients Plots
Description
The plot_integrated_gradients() function replicates the SHAP visualization structure for integrated gradient values, providing the same four graphical modalities adapted to this specific interpretability methodology for neural networks. This function is particularly valuable for understanding feature importance in deep learning architectures where gradients provide direct information about model sensitivity.
Usage
plot_integrated_gradients(analysis_object, show_table = FALSE)
Arguments
analysis_object |
Fitted analysis_object with 'sensitivity_analysis(methods = "Integrated Gradients")'. |
show_table |
Boolean. Whether to print Integrated Gradients summarized results table. |
Value
analysis_object
Examples
# Note: For obtaining the Integrated Gradients plot the user needs to
# complete till sensitivity_analysis( ) function of the MLwrap pipeline
# using the Integrated Gradients method.
if (requireNamespace("torch", quietly = TRUE)) {
wrap_object <- preprocessing(df = sim_data,
formula = psych_well ~ depression + emot_intel + resilience,
task = "regression")
wrap_object <- build_model(wrap_object, "Neural Network")
wrap_object <- fine_tuning(wrap_object, "Bayesian Optimization")
wrap_object <- sensitivity_analysis(wrap_object, methods = "Integrated Gradients")
# And then, you can obtain the Integrated Gradients plot.
plot_integrated_gradients(wrap_object)
}
[Package MLwrap version 0.1.0 Index]