Recently I’m interested in quantum computing, it’s a fantastic quantum world. We’re at the start of a new stage in the information revolution.

If quantum physics sounds challenging to you, you are not alone. All of our intuitions are based on day-to-day experiences and are defined by classical physics — so most of us find the concepts in quantum physics counterintuitive at first. In order to comprehend the quantum world, you must let go of your beliefs about our physical world, and develop an intuition for a completely different (and often surprising) set of laws.

Here I translate the quantum computing documentation from the IBM Q Experience¹ to chinese, on the one hand it’s convenient for others to capture the tough concepts, on the other hand it also gives me a solid understanding.

---

I use Asynchronous Advantage Actor Critic (A3C), one of the most popular reinforcement learning method, combined with Monte Carlo Tree Search (MCTS)¹, to find out the RNA molecule folding path of secondary structure.

The neural network which we used is a simple three convolutional layer with two output head that give current RNA state energy and the probabilities of next pairing action.

The fellow diagram illustrates the whole pipeline.

After learning 1000 folding path episodes of one small RNA (30nt), I find many succeed folding paths to the true native secondary structure. Through analyzing those paths we can understand how RNA fold to the native structure (shown as the figure).²

---

This is my some notes of reinforcement learning, the codes of many methods and trained model for playing different Atari games.
Here sharing them all to you for reinforcing your learning of reinforcement learning.

Table of contents

---

• Q-Learning
  • Frozen Lake Game
  • Tic Tac Toe
  • Taxi v2
• Deep Q-Learning Network (DQN)
  • Doom Game
  • Atari Space Invaders
• olicy Gradients (PG)
  • CartPole Game
  • Doom Deathmatch
• Advantage Actor Critic (A2C)
• Asynchronous Advantage Actor Critic (A3C)
• Proximal Policy Optimization (PPO)
• AlphaGoZero Introduction
• Monte Carlo Tree Search (MCTS)
  • Gomoku
• AlphaGomoku

---