This project has two components. Your tean (2-3 members) will create:
Each of these is described below.
Your agent/world program simulates a world or agent of your choice. It's core will be a function that you create that maps States and Actions to Rewards. This function can be a deterministic mapping, or, preferably a probawill bilistic mapping.
The abstract representation you will use is the following:
We will constrain everyone to use the same abstract representation so that learning agents will be able to interact with all of the simulators created by the class.
Part of your job is to come up with an interpretation for the states, actions and reward in your world. Possibilities include:
name:
Boring AgentState:
mood: happy, sad, mad, bored
physical: hungry, sleepy
personality: optimist, pessimist
Action:
State x Action ->
Reward:
name: Agent-with-a-Personality
Action:
State x Action ->
Reward:
State x Action ->
State
name: Robot Agent
State:
location
Action:
State
x Action -> Reward:
State
x Action -> State
However, the actions should be generated uniformly at random, so that the learning agent gets a chance see the rewards for different actions.
Calls Agent Program
to get a training set
Learns
a function
Calls Agent Program
to get an evaluation set
Computes
Optimal set of actions
Calls Agent Program
to evaluate the set of actions
You can use whatever representation you like for the learning agent: neural networks, decisions trees, nearest neighbor, etc.,
Thursday, Dec. 4
Friday, Dec. 5
Midnight Friday, Dec. 12 (Updated with Extension)
7PM Wednesday, Dec. 17 (NO EXTENSIONS)
