readme.md

Safe RL

---

A Comprehensive Survey on Safe Reinforcement Learning

Safe Model-based Reinforcement Learning with Stability Guarantees

Paper Intro Video

A Lyapunov-based Approach to Safe Reinforcement Learning

DeepMind & Facebook AI

NIPS 2018

Blog

Safe Controller Optimization for Quadrotors with Gaussian Processes

ICRA 2017

Citation 153

Talk

The paper propose Modified SAFEOPT.

Safe Exploration for Interactive Machine Learning

Safe Exploration for Optimization with Gaussian Processes

The paper propose SAFEOPT.

The Lyapunov Neural Network: Adaptive Stability Certification for Safe Learning of Dynamical Systems

Reference

• Risk and robustness in RL: Nothing ventured, nothing gained - Shie Mannor

  A talk about the trade-off between safety and robustness.

• Tutorial on Safe Exploration for Reinforcement Learning

• Andreas Krause (ETH Zuerich): "Safe Exploration in Reinforcement Learning"

  A talk about how to explore safely with RL.

• Towards Structural Risk Minimization for RL - Emma Brunskill

  A talk about risk-sensitive RL.

• Lecture by Felix Berkenkamp (CS 159 Spring 2020)

  Safe Exploration for Reinforcement Learning

• Bayesian Optimization & Algorithm Configuration

  Bayesian optimization of CS159

• Machine Learning by mathematicalmonk

  Give an intuitive explaination of the math often used in ML.

• UAI2018 2. Bayesian Optimization

• Gaussian processes by Martin Krasser

• Bayesian optimization by Martin Krasser

• Coursera: Bayesian optimization

• StackExchange: Intuitive Understanding of Expected Improvement for Gaussian Process

• ML Tutorial: Gaussian Processes (Richard Turner)

• 10 Gaussian Processes, pt 1/3 Basics

• A Visual Exploration of Gaussian Processes

• Bzarg: How a Kalman filter works, in pictures

  Intuitively explain Kalman Filter with picture & examples.

• 图说卡尔曼滤波，一份通俗易懂的教程

  Intuitively explain Kalman Filter with picture & examples. The article is translated from Bzarg: How a Kalman filter works, in pictures.