DL-and-RL

Self-Study Guide for Deep Learning and Reinforcement Learning (I am building this guide as I study)

Deep Learning

MOOC/Websites

• Courseara Andrew Ng Deep Learning Specialization [link]
• Convolutional Neural Networks for Visual Recognition [CS2321n]
• Understanding LSTM Neworks [RNN]
• Tensorflow Tutorials [Hvass]
• A Tutorial on 3D Deep Learning [link]
• 3D Deep Learning Workshop [link]
• Machine Learning with Google Cloud Platform [link]]

Book

• Deep Learning, Ian Goodfellow, Yshua Bengio and Aaron Courville [book]

Papers

• Understanding deep learning requires rethinking generalization [Zhang et al]
• Attention is all you need [Vaswani et al.]
• Faster R-CNN [RPN]

Convolutional Neural Network Architecture

• LeNet
• AlexNet
• VGG
• ResNet
• Network in Network [paper]
• Inception Network

Distributed Network for Deep Learning

• DistBelief
• Revisiting distributed synchronous SGD [paper]

Computer Vision: Object Detection

• Retinanet
• YOLO
• SSD
• Faster RCNN
• Mask RCNN

3D Deep Learning

• Frustum PointNets for 3D Object Detection from RGB-D Data [link]

Reinforcement Learning

MOOC/Websites

• Udacity Reinforcement Learning [link]
• Temporal difference learning lecture
• BURLAP Tutorial

Book

• Richard Sutton and Andrew Barto, Reinforcement Learning: An Introduction (2nd Edition Draft, 2017) [Book]

Papers

• Temporal difference learning Sutton 1988
• Q Learning
• Adaptive Heuristic Critic Method (AHC)

Robotics/Motor Skills

• RL of motor skills with policy gradient paper
• Efficient distributed RL through agreement Varshavskaya et al

Multi-Agent Reinforcement Learning

• Survey on Multi-agent RL [Busoniu et al]

Distributed & Scalable Systems

• Ray RLlib: A Composable and Scalable Reinforcement Learning Library [Liang et al]

Deep Reinforcement Learning

MOOC/Website

• Deep Reinforcement Learning, UC Berkeley [CS294]
• Deep RL Bootcamp, UCBerkeley & OpenAI [link]
• Deep learning and reinforcement learning summer school [lectures]
• Deep Reinforcement Learning (John Schulman, OpenAI) [Video]
• Tensorflow Tutorial #16 Reinforcement Learning [link]

Papers

Video Games

• Neural Fitted Q Iteration [NFQ]
• Deep Q Network [DQN][Nature]
• Deep Q Learning [link]
• Deterministic Deep Policy Gradient [DDPG]
• Universal Value function Approximators [UVFA]

Robotics

• End-to-end training of deep visuomotor policies [Levin et al.]
• Hindsight Experience Replay [HER]
• Sim-to-real transfer for robotic control with dynamics randomization [paper]
• Domain randomization for sim-to-real transfer [Tobin et al]
• Vision-based Multi-task Manipulation with Inexpensive Robot [Rahmatizadeh et al]
• DART: Noise Injection for Robust Imitation Learning [Laskey et al]

Surgical Robotics

• Multilateral Surgical Pattern Cutting with DRL [link]
• (Unsupervised Learning) Transition State Clustering [link]
• Learning by Obsercation for Surgical Subtask [Murali et al]

Scalable & Distributed Systems

• Parrellel Methods for DRL [DeepMind Paper]
• DDRL through Agreement [paper]
• Distributed Deep Q-Learning
• Deep multi-user RL [paper]
• Massively parallel methods for DRL [paper]
• HORDE: A scalable real-time architecture for learning knowledge from unsupervised sensorimotor interaction []

Model-based deep reinforcement learning

• Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning

Inverse Reinforcement Learning/Inverse Optimal Control [Nagabandi et al.]

• Guided Cost Learning: Deep inverse optimal control via policy opimization [Finn et al]
• Generative Adversarial Imitation Learning Ho and Ermon code

Code

• openAI baselines
• Neural Network Dynamics for Model-based deep reinforcement learning with model-free fine-tuning code
• End-to-end learning with deep visuomotor policies code
• Vision-based multi task manipluation with inexpensive robot code