Interpretability

• Interpretable Machine Learning (A Guide for Making Black Box Models Explainable Christoph Molnar (book) link

• sparse networks with relatively small amount of neurons
  • biology
  • Ordinary Neural Circuits
    • RL-based architecture learning algorithm
    • article link
• CNNs: visualbackprop: efficient visualization link

Causal Learning

• Elements of Causal Inference Jonas Peters (book) link
• Causal modeling (article) link

• Causal inference in Statistics (article) link

• graphical models
• applying Bayesian methods link
  • to decompose the learning problem into a set of Bayesian inference sub-problems
  • effective in time-series setting link
• Deep Kalman filters article
  • do not seek to learn the true underlying causal graph structure but rather seek to use do-calculus to observe the effect of interventions under a causal interpretation
• continuous-time Bayesian networks link
• learn ODEs
  • under certain conditions imply a structured causal model link
  • Liquid Time-constant Neural Networks
• CausalVAE