Using Hierarchical Reinforcement Learning to Solve a Problem with Multiple Conflicting Sub-problems.

Slides:



Advertisements
Similar presentations
Managing Stress How should you start dealing with stress?
Advertisements

Scientists group and classify things in our world – just like we can group shapes together. How do you think a scientist would say these things are alike?
Chemical Analysis of Water
Reinforcement Learning
Mrs. Burkhart 2 nd Grade. Everyone has wants and needs. They can be different for each person. So what is the difference between a want and need?
Note for C 1.1: Single-celled organisms have all the characteristics of living things.
TOPIC 12 CREATING CLASSES PART 1 1 Notes adapted from Introduction to Computing and Programming with Java: A Multimedia Approach by M. Guzdial and B. Ericson,
Read each slide. Look at each picture. Answer each question.
Interactions Among Living Things
Brooke Ard 5 th Grade Math/Science Griggs Road Elementary.
Review of learning theory and evolutionary theory.
10 Wilderness Survival Lessons From Hatchet. 1. Take Inventory of Your Supplies “It kept coming back to that. He had nothing. Well, almost nothing. As.
Theories of Motivation
Animal Welfare During Animal Health Emergencies. Animal Welfare ●Ethical responsibility ●Ensuring animal well being ●Physical and mental ●Consideration.
LEARNING FROM OBSERVATIONS Yılmaz KILIÇASLAN. Definition Learning takes place as the agent observes its interactions with the world and its own decision-making.
Hierarchical Reinforcement Learning Ersin Basaran 19/03/2005.
LEARNING FROM OBSERVATIONS Yılmaz KILIÇASLAN. Definition Learning takes place as the agent observes its interactions with the world and its own decision-making.
Myers’ EXPLORING PSYCHOLOGY (6th Ed) Chapter 10 Motivation Modified from: James A. McCubbin, PhD Clemson University Worth Publishers.
Chapter 1: An Overview of Computers and Programming Languages
8.L.5.1 Summarize how food provides the energy and the molecules required for building materials, growth and survival of all organisms (to include plants)
Reinforcement Learning and Soar Shelley Nason. Reinforcement Learning Reinforcement learning: Learning how to act so as to maximize the expected cumulative.
Basic needs of living things
Lesson 1 Your Mental and Emotional Health Do you think you have a positive outlook? Do you look forward to facing life’s challenges?
Problem of the Day Problem of the Day Multiplication next.
Chapter 1 Earth: The Water Planet Section 1 How Is Water Important?
Project Title Goes Here Names of Team Members (and Addresses if you would like to include) 1.
lesson 5.1 THE IPDE PROCESS
1 PLAN RECOGNITION & USER INTERFACES Sony Jacob March 4 th, 2005.
Test-Taking Strategies
What are the perfect conditions for growing a bean plant? How can we find out?
Biological and Social Motives
1 CSC 8520 Spring Paula Matuszek Kinds of Machine Learning Machine learning techniques can be grouped into several categories, in several ways: –What.
Your Mental and Emotional Health Mental/Emotional Health – the ability to accept yourself and others, adapt to and manage emotions, and deal with the demands.
Christen Cloud.
Using Hierarchical Reinforcement Learning to Balance Conflicting Sub- problems By: Stephen Robertson Supervisor: Phil Sterne.
People who are starving see food as a means of survival.
Order of Operations - rules for arithmetic and algebra that describe what sequence to follow to evaluate an expression involving more than one operation.
Curiosity-Driven Exploration with Planning Trajectories Tyler Streeter PhD Student, Human Computer Interaction Iowa State University
Ecosystems. What is an Ecosystem?  An ecosystem is a plant and animal community made up of living and nonliving things that interact with each other.
Harry Harlow And Abraham Maslow
Functions of Management
Native Plants and Animals: Introduction to Habitats.
Living Things and the Environment EQ: How do I define the various levels of an ecosystem?
Theories of Motivation
The Data Handling Cycle. Specifying the problem and Planning (Posing questions) Does our class have a healthy diet? I need to ask everyone what they eat.
Finding Optimal Solutions to Cooperative Pathfinding Problems Trevor Standley Computer Science Department University of California, Los Angeles
Characteristics of Living Things Unit 2. It’s Alive! Or Is It?
What is life? Mr. Morris’s 1 st Grade Science Class.
Would You Survive In The Wilderness?. Put the following into the order in which you would need to do them to have the best chance of survival. Build a.
Welcome to Mr. Crocco’s Living Environment. What is Science? Science is an organized way of using evidence to learn about the real or natural world. The.
人教修订版 高中二年级 ( 上 ) Unit 6. Listening Revision My dream 1. I have a dream. My dream is to become a cartoonist in the future. Then I can draw fine pictures.
Ecosystems 5.L.4B.4 Construct scientific arguments to explain how limiting factors (including food, water, space, and shelter) or a newly introduced.
Warm up: What are the 6 characteristics of all living things?
MOTIVATION B.A.-I P.G.G.C.G-42 Chandigarh By: Ms. Amandeep.
I AM the Bread of Life John 6
Mission to Mars STEM/Robotics Project
What resources do all humans need in order to survive on Earth
CMSC 471 Fall 2009 RL using Dynamic Programming
Chapter 4: Dynamic Programming
Chapter 4: Dynamic Programming
CIS 488/588 Bruce R. Maxim UM-Dearborn
I Am The Bread of Life I am the bread of life
Chapter 19; Natural Systems and Interactions
“I am the Bread of Life” – Andrew Chinn
Zonesa Kerley & Taylor Engstrom
Chapter 4: Dynamic Programming
An environment is made up of all the living and non-living things with which an organism (living thing) may interact.
Survival Skills.
lesson 5.1 THE IPDE PROCESS
By: Stephen Robertson Supervisor: Phil Sterne
Presentation transcript:

Using Hierarchical Reinforcement Learning to Solve a Problem with Multiple Conflicting Sub-problems

Reinforcement Learning Involves an agent interacting with an environment The agent can be in one of various states in the environment The agent is not told which action is correct, but is given a measure of an action for a given state After a while the agent develops a policy

The curse As complexity of the environment grows, state space increases exponentially We can try to cleverly reduce state space Hierarchical reinforcement learning

Hierarchical Reinforcement Learning A complex problem can often be broken up into multiple conflicting sub-problems Hierarchical reinforcement learning can handle this Deals with each sub-problem separately using reinforcement learning Decides which sub-problem to attempt next using reinforcement learning

A Practical Example: The Mars Lander Perform Various Conflicting Tasks: Explore the terrain Collect soil samples Return to base for refuelling

My Project Apply hierarchical reinforcement learning to a complex problem Consist of an agent existing in an environment where it will have to achieve an overall goal Agent will be a primitive creature trying to survive in the wilderness

My Project The overall goal will be for the creature to remain happy or comfortable in the wilderness Overall goal can be divided into sub-goals These sub-goals will be: –Eating food –Drinking water –Resting under a Shelter –Repairing Shelter –Avoiding hazards

The Gridworld

Motivation for this approach X pos Y pos Hunger Thirst Fatigue Shelter Condition 13 x 13 x 10 x 10 x 10 x 10 = Possible states Sub-goals separated out: (Xpos, Ypos, hunger), (Xpos, Ypos, Thirst) (Xpos, Ypos, Fatigue), (Xpos, Ypos, Shelter Condition) (13 x 13 x 10) x 4 =1690 x 4 = 6760 Possible states