Markdown Research Notes 2019

#Machine Learning #Bioinformatics #Reinforcement Learning

---

Contents:

• RNA Secondary Structures (1/7)
  • 1. Secondary Structure Graphs
    • 1.1 Introduction
    • 1.2 Secondary Structure Graphs
    • 1.3 Mountain Plots and Dot Plots
    • 1.4 Trees and Forests
    • 1.5 Notes.
• RNA Secondary Structures (2/7)
  • 2. Loop-Based Energy Model
    • 2.1 Loop decomposition
    • 2.2 Energy Parameters
    • 2.3 Notes
• RNA Secondary Structures (3/7)
  • 3. The RNA Folding Problem
    • 3.1 Counting structures and maximizing base pairs
    • 3.2 Backtracing(or Backtracking)
    • 3.3 Energy minimization in the loop-based energy model
• Markov Decision Process For Predict RNA Secondary Structure #1
  • 1. Introduction
    • 1.1 Markov Decision Process(MDP)
    • 1.2 Reinforcement Learning(RL)
    • 1.3 Reinforcement Learning with Example
    • 1.4 RNA Secondary Structure Prediction
  • References
• Markov Decision Process For Predict RNA Secondary Structure #2
  • 2. RNA Secondary Structure Prediction in Reinforcement Learning
    • 2.1 introduction
    • 2.2 Reinforcement Learning with RNA Secondary Structure Prediction
    • 2.3 Reinforcement Algorithm with Deep learning
  • References
• About the Markov Decision Process(MDP)
  • 1. introduction
  • 2. Markov Process(MP)
  • 3. Markov Reward Process(MRP)
  • 4. Markov Decision Process(MDP)
  • References
• A Brief Review of Probability Theory
  • Probability
    • 1. introduction
    • 2. A brief review of probability theory
  • References
• One of the RL(Reinforcement Learning) method - Cross Entropy
  • 1. Introduction
  • 2. Taxonomy of RL methods
    • Model-Free
    • Model-Based
    • Policy-Based
    • Value-Based
    • On-Policy
    • Off-Policy
    • Cross-entropy
  • 3. Practical cross-entropy
  • 4. Summary
  • References
• The Bellman equation of optimality
  • Bellman equation with deterministic example
  • Bellman equation with stochastic example
  • Conclusion
  • Reference
• Why we use the Q-function?
  • Value of state & Value of action
  • Q-function with simple example
  • References
• The value iteration method
  • Value iteration with Example
  • Value iteration algorithm
    • A. Value of states
    • B. Value of Actions(Q)
  • Limits of value iteration methods
    • Size of state space problem
    • State transition probability problem
  • References
• Deep Q-Networks
  • Real-life value iteration
  • Tabular Q-learning
  • Deep Q-learning
  • Interaction with the environment
    • Epsilon-Greedy method
  • Replay Buffer
  • References
• Policy Gradients(PG)
  • Value and Policy
    • Review with respect to Q-learning
    • Policy in Q-learning
  • Why Policy?
  • Policy Representation
  • Policy Gradient
  • The REINFORCE method
    • REINFORCE
    • Difference with Q-learning
  • Policy-Based versus Value-Based methods
  • REINFORCE issues
    • Too Long Episode
    • High Gradient Variance
    • Exploration - Entropy Bonus
  • References
• Graph Structure’s Similarity methods
  • Default planar graphs for similarity test
  • Euclidean Similarity
  • Cosine Similarity
  • Eigenvalues Similarity
  • Results
  • References