Learning to Plan

Photo by Artyom Kabajev on Unsplash
Schema of an agent acting in an environment and receiving a reward
Figure 1: Agent acting in the environment and receiving a reward.
Figure 2: (left) an agent deployed in the battlefield; (right) possible actions for the agent.
Figure 2: (left) an agent deployed in the battlefield; (right) possible actions for the agent.
  1. Move north to secure the objective (represented by a green arrow);
  2. Left cover and engage the unit at the north (orange arrow);
  3. Left cover and engage the lonely unit at the south (blue arrow); and
  4. Do nothing.
  • Agent: I am in a position at X,Y.
  • Sergeant: Are there enemies near you?
  • Agent: I can sense that there are two enemy units near.
  • Sergeant: Is there an objective near?
  • Agent: There is one objective north of my position, clear of enemies.
  • Sergeant: And what have you decided to do?
  • Agent: I decided to do nothing (4th option).
  • Sergeant: . . .
  • The agent is situated in a state (s), and the agent chooses an action (a): to do nothing
  • The agent executes the selected action and modifies the current situation, by entering another state (s’).
  • The agent obtains a reward (r(s’)), according to the outcome of the chosen action (a).
  • Start again.
  • Agent: I am in position…bla bla…
  • …[REDACTED CONVERSATION]…
  • Agent: I moved north and secure the objective
  • Sergeant: Good!
Vindicare assasin from Warhammer 40k.
Vindicare Assassin from Warhammer 40k. In the far future it is better not to approach figures that are sitting alone.

--

--

--

I am a Über nerd, interested in Robotics, Machine Learning and Computer Science in general, as well as Entrepreneurship.

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Using image data, predict the gender and age range of an individual in Python

Vascular Clog Loss Classification: An Advanced Alzheimer’s Research Using ConvNets

A Threshold Approach for Dynamic Skin Segmentation

Parkinson’s disease detection — on cAInvas

Class activation maps: Visualizing neural network decision-making

Machine learning with Decision Trees

85e106ae-bc60-4759-8bf5-70fe3c126fd5-original

Supervised Machine Learning (Regression Algorithms)

Machine Learning for HR: 5 Steps to Corporate Chatbot Development

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Ignasi Andres

Ignasi Andres

I am a Über nerd, interested in Robotics, Machine Learning and Computer Science in general, as well as Entrepreneurship.

More from Medium

Reinforcement Learning -The Basics

AI Talks about Traveling

INTRODUCTION TO REINFORCEMENT LEARNING AND TYPES

How to make money with machine learning: value betting on predicted UFC fight outcomes