cyberneticlibrary

Explain ML model predictions

shapskillsetup L227,559
K-Dense-AI/scientific-agent-skills
What it does

Explain ML model predictions via Shapley values

Best for

Debugging model decisions and checking for fairness/bias before production deployment.

Inputs
  • · trained model
  • · test data
Outputs
  • · SHAP values
  • · feature importance plots
  • · waterfall/force visualizations
Requires
  • · SHAP library
  • · tree-based/deep learning models
Preconditions

Model must be trained; input data compatible with model

Failure modes

Model type unsupported (use KernelExplainer fallback); memory exceeded on large datasets

Trust signals
  • · MIT license
  • · K-Dense Inc. skill-author
  • · supports tree/neural/linear/black-box models