reward
41
5.3
- noun
- /rɪˈwɔrd/
- Specialized
:
recompensa
- The agent receives a positive reward for successfully completing its task in the simulation.
- extrinsic reward
- reward structure
- reward system
So standard rewards are numbers and numbers are hard to come by.
- So standard rewards are numbers and numbers are hard to come by.
- Why is it useful to combine the reward functions?
- So you're you're you're maximizing reward within these constraints.
Examples
-
Adjustments in the parameters led to a more effective reward system for the learning algorithm.
-
Basically, you have a built in reward system in your brain.
-
Go on holiday, try and find reward in smaller things.
-
I think understanding the world is its own reward in a lot of ways.
-
You want the reward without putting in the work.
-
"You have got to do things for other people and not expect any kind of reward."
-
The job comes with more responsibility, but it offers a lot more rewards.
-
These simple daily investments have giant rewards and they take very little time.
-
Recent research has questioned the basic assumptions upon which these rewards are based.
Academic text (1992) -
It was risky, but the possible reward was also high.
Synonyms
439
3.9
A thing that makes a person want to do something, often a reward
31
3.8
A good or bad result that follows an action
7
6.6
An inner goal that guides a robot's actions
How Positive
Surface Forms
Etymology
The machine learning use of reward keeps the old idea of 'something given back' (think re- 'back' and ward 'toward'), but as a number or score. So a reward in AI is the value given to an agent to show how good a particular action or state was.