Lecture 2 Additional Software Tools

• Python:
  • allennlp
  • alibi
  • anchor
  • DiCE
  • interpret
  • lime
  • shap
  • PiML-Toolbox
  • tf-explain
• R:
  • ALEPlot
  • DALEX
  • ICEbox
  • iml
  • Model Oriented
  • pdp
  • shapFlex
  • vip
• Python, R or other:
  • h2o-3

Lecture 2 Additional Software Examples

• Global and Local Explanations of a Constrained Model
• Building from Penalized GLM to Monotonic GBM
• Monotonic XGBoost models, partial dependence, individual conditional expectation plots, and Shapley explanations
• Decision tree surrogates, LOCO, and ensembles of explanations
• Machine Learning for High-risk Applications: Use Cases (Chapter 6)

Lecture 2 Additional Reading

• Introduction and Background:
  • On the Art and Science of Explainable Machine Learning
  • Proposed Guidelines for the Responsible Use of Explainable Machine Learning
• Post-hoc Explanation Techniques:
  • A Unified Approach to Interpreting Model Predictions
  • Anchors: High-Precision Model-Agnostic Explanations
  • Elements of Statistical Learning - Section 10.13
  • Extracting Tree-Structured Representations of Trained Networks
  • Interpretability via Model Extraction
  • Interpretable Machine Learning - Chapter 6 and Chapter 7
  • Peeking Inside the Black Box: Visualizing Statistical Learning with Plots of Individual Conditional Expectation
  • Towards Better Understanding of Gradient-based Attribution Methods for Deep Neural Networks
  • Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models
  • “Why Should I Trust You?” Explaining the Predictions of Any Classifier
• Problems with Post-hoc Explanation:
  • General Pitfalls of Model-Agnostic Interpretation Methods
  • Limitations of Interpretable Machine Learning Methods
  • When Not to Trust Your Explanations