README.md

reading-list

This is a collection of interesting papers that I have read so far or want to read. Note that the list is not up-to-date.

Table of Contents

1. General Deep Learning
2. Conformal Prediction
3. Differential Geometry in Deep Learning
4. Dimensionality Reduction
5. Thompson Sampling
6. Deep Reinforcement Learning
7. Reinforcement Learning
8. Bandit algorithms
9. Optimization
10. Statistics
11. Probability modeling and inference
12. Books, courses and lecture notes
13. Blogs and tutorial
14. Schools

1. General deep learning

• 2023, Your diffusion model secretly knows the dimension of the data manifold
• 2022, Regularising Inverse Problems with Generative Machine Learning Models
• 2021, SCORE-BASED GENERATIVE MODELING THROUGH STOCHASTIC DIFFERENTIAL EQUATIONS
• 2021, Distilling Robust and Non-Robust Features in Adversarial Examples by Information Bottleneck
• 2021, Slot Machines: Discovering Winning Combinations of Random Weights in Neural Networks
• 2021, Loss landscapes and optimization in over-parameterized non-linear systems and neural networks
• 2021, Why flatness correlates with generalization for Deep NN
• 2021, The Modern Mathematics of Deep Learning
• 2021, The Principles of Deep Learning Theory
• 2020, Neural tangent kernel
• 2018, Lipschitz regularity of deep neural networks: analysis and efficient estimation
• 2015, Weight Uncertainty in Neural Networks
• 1998, LeCun, Efficient BackProp

2. Conformal prediction

• 2022, Conformal Prediction: a Unified Reviewof Theory and New Challenges
• 2022, Conformal Off-Policy Prediction in Contextual Bandits
• 2020, Conformal Prediction Under Covariate Shift
• 2019, Conformalized Quantile Regression
• 2005, Algorithmic Learning in a Random World

3. Differential geometry in deep learning

• 2020, Neural Ordinary Differential Equations on Manifolds
• 2019, Efficient Approximation of Deep ReLU Networks for Functions on Low Dimensional Manifolds
• 2019, Diffeomorphic Learning
• 2019, Deep ReLU network approximation of functions on a manifold
• 2019, Efficient Approximation of Deep ReLU Networks for Functions on Low Dimensional Manifolds
• 2016, Deep nets for local manifold learning

4. Dimensionality reduction

• 2020, Stochastic Neighbor Embedding with Gaussian and Student-t Distributions: Tutorial and Survey
• 2015, Parametric nonlinear dimensionality reduction using kernel t-SNE
• 2009, Learning a Parametric Embedding by Preserving Local Structure

5. Thompson sampling

• 2020, A Tutorial on Thompson Sampling
• 2020, Neural Thompson Sampling
• 2018, Deep Contextual Multi-armed Bandits

6. Deep Reinforcement Learning

• Famous applications

  • 2022, CICERO

• Algorithms

  • List of algorithms
  • DQN Pop art
  • 2018, Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
  • 2017, Reinforcement Learning with Deep Energy-Based Policies

• DQN Convergence

  • 2020, A theoretic analysis of DQN
  • 2020, A Finite-Time Analysis of Q-Learning with Neural Network Function Approximation
  • 2019, Towards Characterizing Divergence in Deep Q-Learning

• Soft Q learning theory

  • 2018, Equivalence Between Policy Gradients and Soft Q-Learning

• Exploration papers

  • 2020, Planning go explore via self-supervised world models
  • 2017, #Exploration: A Study of Count-Based Exploration for Deep Reinforcement
  • 2016, Unifying Count-Based Exploration and Intrinsic Motivation
  • 2022, Exploring through Random Curiosity with General Value Functions

• Stability

  • 2021, Enforcing Robust Control Guarantees with neural network polciies
  • 2018, Control-Theoretic Analysis of Smoothness for Stability-Certified Reinforcement Learning

• Dreamer algorithm

  • 2021, Mastering Atari with discrete world-models
  • 2020, Dream to control
  • 2020, Planning to explore via supervised world-models
  • 2019, Learning latent dynamics from pixels

• Goal conditioned RL

  • 2021, Adversarial Intrinsic Motivation for Reinforcement learning

• Distributional RL

  • 2021, GMAC: A Distributional Perspective on Actor-Critic Framework
  • 2020, SAMPLE-BASED DISTRIBUTIONAL POLICY GRADIENT
  • 2019, Statistics and Samples in Distributional Reinforcement Learning
  • 2018, DISTRIBUTED DISTRIBUTIONAL DETERMINISTIC POLICY GRADIENTS

7. Reinforcement Learning

• Papers

  • 2023, Empirical Design in Reinforcement Learning
  • 2023, An Analysis of Quantile Temporal-Difference Learning
  • 2022, Understanding Policy Gradient Algorithms: A Sensitivity-Based Approach
  • 2021, Sample Complexity of Asynchronous Q-Learning: Sharper Analysis and Variance Reduction
  • 2021, Adaptive Sampling for Best Policy Identification in MDPs
  • 2020, Provably Efficient Exploration for Reinforcement Learning Using Unsupervised Learning
  • 2019, Revisiting the Softmax Bellman Operator: New Benefits and New Perspective
  • 2019, Q-learning with UCB Exploration is Sample Efficient for Infinite-Horizon MDP
  • 2019, Provably Efficient Reinforcement Learning with Linear Function Approximation
  • 2018, Deep Reinforcement Learning that Matters
  • 2018, Is Q-learning Provably Efficient?
  • Adaptive sampling for policy identification
  • On Function Approximation in Reinforcement Learning: Optimism in the Face of Large State Spaces
  • 2016, Learning the Variance of the Reward-To-Go
  • 2012, Policy Gradients with Variance Related Risk Criteria
  • 2009, An Analysis of Reinforcement Learning with Function Approximation
  • 2008, An analysis of model-based Interval Estimation for Markov Decision Processes
  • 2006, PAC Model-Free Reinforcement Learning
  • 2004, Bias and Variance in Value Function Estimation
  • 2001, Convergence of Optimistic and Incremental Q-Learning
  • 2001, TD Algorithm for the Variance of Return and Mean-Variance Reinforcement Learning
  • 2000, Convergence Results for Single-Step On-Policy Reinforcement-Learning Algorithms
  • 1993, Convergence of Stochastic Iterative Dynamic Programming Algorithms
  • 1992, Reinforcement learning applied to linear quadratic regulation
  • 1982,The Variance of Discounted Markov Decision Processes

• Zero shot RL

  • 2023, Does Zero-Shot Reinforcement Learning Exist?
  • 2021, Learning One Representation to Optimize All Rewards
  • 2017, Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning

• Constrained/Safe RL

  • 2022, Safety-constrained reinforcement learning with a distributional safety critic
  • 2022,Constrained Variational Policy Optimization for Safe Reinforcement Learning
  • 2022, TRC: Trust Region Conditional Value at Risk for Safe Reinforcement Learning
  • 2022, Towards Safe Reinforcement Learning via Constraining Conditional Value-at-Risk
  • 2022, SAAC: Safe Reinforcement Learning as an Adversarial Game of Actor-Critics
  • 2019, Benchmarking Safe Exploration in Deep Reinforcement Learning
  • 2017, Constrained Policy Optimization
  • 2017, Risk-Constrained Reinforcement Learning with Percentile Risk Criteria
  • 2015, Variance-Constrained Actor-Critic Algorithms for Discounted and Average Reward MDPs
  • 2015, A Comprehensive Survey on Safe Reinforcement Learning
  • 2015, Risk-Sensitive and Robust Decision-Making: a CVaR Optimization Approach

• Off policy evaluation

  • 2022, A Review of Off-Policy Evaluation in Reinforcement Learning
  • 2022, Conformal Off-Policy Prediction in Contextual Bandits
  • 2020, CoinDICE: Off-Policy Confidence Interval Estimation
  • 2018, Breaking the Curse of Horizon: Infinite-Horizon Off-Policy Estimation
  • 2015, High Confidence Policy Improvement
  • 2015, High Confidence Off-Policy Evaluation
  • 2000, Eligibility Traces for Off-Policy Policy Evaluation

8. Bandit algorithms

• Papers

  • 2023, Quantile Bandits for Best Arms Identification
  • 2020, Neural Contextual Bandits with Deep Representation and Shallow Exploration
  • 2020, Neural Contextual Bandits with UCB-based Exploration
  • 2016, Optimal Best Arm Identification with Fixed Confidence
  • 2016, On the Complexity of Best-Arm Identification in Multi-Armed Bandit Models
  • 2016, Explore First, Exploit Next: The True Shape of Regret in Bandit Problems
  • 2011, Online Least Squares Estimation with Self-Normalized Processes: An Application to Bandit Problems
  • 2002, Finite-time Analysis of the Multiarmed Bandit Problem
  • 2002, Using Confidence Bounds for Exploitation-Exploration Trade-offs
  • 2002, THE NONSTOCHASTIC MULTIARMED BANDIT PROBLEM∗

9. Optimization

Min-max optimization

• 2021, A mean-field analysis of two-player zero-sum games
• 2021, The Limits of Min-Max Optimization Algorithms: Convergence to Spurious Non-Critical Sets
• 2020, On the Convergence of Single-Call Stochastic Extra-Gradient Methods
• 2020, Non-convex Min-Max Optimization: Applications, Challenges, and Recent Theoretical Advances
• 2020, On Gradient Descent Ascent for Nonconvex-Concave Minimax Problems
• 2020, Robust Reinforcement Learning via Adversarial training with Langevin Dynamics
• 2018, Finding Mixed Nash Equilibria of Generative Adversarial Networks

10. Statistics

• 2022, A short note on an inequality between KL and TV
• 2020, A Tutorial on Quantile Estimation via Monte Carlo
• 2012, CONCENTRATION INEQUALITIES FOR ORDER STATISTICS
• 1996, IMPORTANCE SAMPLING FOR MONTE CARLO ESTIMATION OF QUANTILES
• 1987, Better Bootstrap Confidence Intervals
• 1982, SOME METHODS FOR TESTING THE HOMOGENEITY OF RAINFALL RECORDS

11. Probability modeling, inference

• 2023, SEMI-IMPLICIT VARIATIONAL INFERENCE VIA SCORE MATCHING
• 2021, Normalizing Flows for Probabilistic Modeling and Inference
• 2020, Improved Techniques for Training Score-Based Generative Models
• 2019, Variational approximations using Fisher divergence
• 2018, Semi-Implicit Variational Inference
• 2018, Variational Inference: A Review for Statisticians
• 2017, Variational Hamiltonian Monte Carlo via Score Matching
• 2013, Stochastic Variational Inference
• 2013, Auto-Encoding Variational Bayes
• 2005, Estimation of Non-Normalized Statistical Models by Score Matching

12. Lecture notes, books and courses

• 2021, Regularization in RL, Google
• CS 6789: Foundations of Reinforcement Learning
• RL Book Theory
• Reinforcement Learning: an introduction
• Bandit algorithms
• 2021, Lecture Notes for Statistics 311/Electrical Engineering 377
• 2015, Rademacher complexities and VC Dimension
• 2013, An introduction to stochastic approximation
• 2006, System identification and the limits of learning from data
• Deep Learning, Goodfellow et al., 2016
• The Elements of Statistical Learning, Hastie, Tibshirani, and Friedman, 2009
• Machine Learning: a Probabilistic Perspective, Murphy, 2012
• Probability Theory: The Logic of Science, E. T. Jaynes, 2003
• CS285 at UC Berkeley, Deep Reinforcement Learning
• CS234 at Stanford University, Reinforcement Learning
• 15.097 at MIT, Prediction: Machine Learning and Statistics
• 2008, Graphical Models, Exponential Families, and Variational Inference, Wainwright and Jordan

13. Blogs

• Deep Reinforcement Learning Doesn't Work Yet
• Distill's publication on Feature Visualization
• Lil'Log, Blog on machine learning

14. Schools

• https://www.math.unipd.it/~vargiolu/home/link.html
• School of mathematics
• Machine learning schools
• Prairie summer school

Papers to add

Deep Bandits Show-Off: Simple and Efficient Exploration with Deep Networks https://proceedings.neurips.cc/paper/2016/file/abd815286ba1007abfbb8415b83ae2cf-Paper.pdf