Machine Learning & Artificial Intelligence · Artificial Intelligence & Deep Learning

Artificial Intelligence: Ethics, Interpretability, and Robustness

Public introduction · full guide in Eikos

Introduction

What this guide covers

Introduction Critical Systems · Security Engineering

Overview

Topics covered

  • Introduction
  • Ethics, Interpretability and Robustness in Artificial Intelligence
  • Introduction to Trustworthy AI and Security Philosophy
  • Mathematical Framework of Robustness and Adversarial Disturbances
  • 2.1 Madry Minimax Formulation
  • 2.2 The Disturbance Space

Related concepts

Main connections

  • Model interpretability (SHAP / LIME)
  • Calculus & differentiation
  • Diagnostic test evaluation & ROC
  • Algorithmic fairness & disparity metrics
  • Mean & expected value
  • Normal (Gaussian) distribution

Compatible notation

Mathematical sample in LaTeX

The expression is rendered accessibly, with its LaTeX source preserved below.

minθ𝔼(x,y)∼𝒟[maxδ∈𝒮ℒ(fθ(x+δ),y)]
View LaTeX source\min_{\theta} \mathbb{E}_{(x, y) \sim \mathcal{D}} \left[ \max_{\delta \in \mathcal{S}} \mathcal{L}(f_\theta(x + \delta), y) \right]