Course:
Advanced Topics in Deep Reinforcement learning
Plan our classes
The idea of this course is to concentrate on modern research in the RL and to analyze significant articles over the past few years.
Date
Lecturer
Description
Links
Rl#1: 13.02.2020
Exploration in RL
Sergey Ivanov
  • Random Network Distillation [1]
  • Intrinsic Curiosity Module [2,3]
  • Episodic Curiosity through Reachability [4]
Rl#2: 20.02.2020
Imitation and Inverse RL
Just Heuristic
  • Imitation Learning[5]
  • Inverse Reinforcement Learning [6,7]
  • Learning from Human Preferences [8]
Rl#3: 27.02.2020
Hierarchical Reinforcement Learning
Petr Kuderov
  • Hierarchical task decomposition
  • Hierarchical learning with predefined sub-goals [9]
  • Automatic sub-goal discovery[11,12]
tba
Rl#4: 5.03.2020
Evolutionary Strategies in RL
tba
  • Genetic Algorithms for Policy Learning[13,14]
  • Combining Evolutionary Strategies with RL[15,16]
  • Improving Exploration with Evolutionary Strategies[17]
tba
Rl#5: 12.03.2020
RL for Combinatorial optimization
tba
  • Why use RL in Combinatorial Optimization?
  • Transformer-based graph encoding[18]
  • Exploratory Combinatorial Optimization with Reinforcement Learning [19]
  • Scaling to Very Large Problems[20]
tba
Rl#6: 19.03.2020
Distributional Reinforcement Learning
tba
  • Estimating Q-value distribution[21]
  • Quantile Regression for Distributional RL[22,23]
  • Expected RL vs Distributional RL[24]
tba
Rl#7: 26.03.2020
Multi-Agent Reinforcement Learning
tba
  • Intra-Agent and Inter-Agent Knowledge Transfer[25]
  • Cooperative and Competitive MARL[26, 27]
  • Learning Inter-Agent Communication[28]
tba
Rl#8: 2.04.2020
Variational Inference in RL
tba
  • Soft Actor-Critic[29,30]
  • Temporal Difference VAE [31]
tba
Rl#9: 9.04.2020
Reinforcement Learning at Scale
tba
  • On-policy Distributed RL: IMPALA[32]
  • Off-policy Distributed RL: Ape-X[33], R2D2[34]
  • Distributed Reinforcement Learning frameworks
tba
Rl#10: 16.04.2020
Memory in RL and POMDP
tba
  • Partially Observable Markov Decision Process (POMDP)
  • Learning recurrent networks in RL: R2R2[35]
  • Memory Augmented Neural Networks [36]
  • Semi-parametric Memory Architectures[37]
tba
Rl#11: 23.04.2020
Model Based RL
tba
  • Planning with Model[38]
  • Model Learning[39]
  • Estimating Uncertainty of the model[40]
  • Combining Model Free and Model Based RL[41]
tba
Rl#12: 30.04.2020
Transfer and Multi-Task Learning
tba
  • Batch Multi-Task Learning[42]
  • Policy Generalization: Generalization to more complex versions of a task[43,44]
  • Task Generalization: Generalization to new goals[45]
  • Sim2Real: Policy transfer from simulation to the real world[46,47]
tba
Rl#13: 7.05.2020
Meta-Reinforcement Learning
tba
  • Inductive bias of a learning algorithm
  • Meta Learning: Learning to Learn[48, 49, 50]
  • Lifelong Learning: Catastrophic Forgetting Problem[51]
tba
Supports

Keep in touch
Sign up to get the latest updates on our community and news.
We promise not to spam nor use your email for anything else.