Reinforcement Learning || Part 04
In my previous articles, I have given a brief detail on reinforcement learning, some of its algorithms, and reinforcement agents.
As the final article on reinforcement learning let’s summarize previous articles and get to know about a few examples of reinforcement learning.
There have main three areas in machine learning which are, Supervised learning, unsupervised learning, and reinforcement learning. In reinforcement learning, it gives solutions for various kinds of planning and control problems. Major components in reinforcement learning are policy, model of the environment, reward signal, and value function.
Here are some of the main algorithms that can be used in reinforcement learning.
- Tabular Methods
- Approximate Solution Methods
- Monte Carlo Method and Temporal-Difference Learning Method
- Policy-Based Reinforcement Learning
- Deep Q-Network Algorithm
Reinforcement agent act in an unknown environment, it learns the optimal behavior to achieve the value function to get the reward. The behavior of the agent will be controlled by the policy. For more information, you can refer to my previous articles.
Reinforcement learning is used in many industries like gaming, health care, traffic control, etc.
In computer games. AlphaGo, ATARI, and Chess are some examples of those computer games. Also, image processing techniques use reinforcement learning. In the healthcare section, to analyze medical images like CT, analyzing traffic images in CCTV cameras for crowd and road processing in real-time. Further, robotics use reinforcement learning to understand surroundings via its visual sensors.
In natural language processing, reinforcement learning is used for text summarization, and question-answering. The agent learns basic language patterns and then uses that knowledge to summarize or provide answers for a new given scenario.
Further, self-driving cars, personalized advertisement recommendation systems, video recommendations systems product recommendations systems also use reinforcement learning to gain their users.
