Explanation-based Learning for Planning: Difference between revisions
Jump to navigation
Jump to search
(Created page with "An Explanation-based Learning for Planning is a planning task based on explanation-based learning. * <B>AKA:</B> Speedup Learning for Planning. * <B>See:</B> [...") |
(→2017) |
||
| Line 7: | Line 7: | ||
=== 2017 === | === 2017 === | ||
* ([[Kambhampati & Yoon, | * ([[Kambhampati & Yoon, 2017]]) ⇒ Kambhampati S., Yoon S. (2017). [https://link.springer.com/referenceworkentry/10.1007/978-1-4899-7687-1_97 "Explanation-Based Learning for Planning"]. In: Sammut C., Webb G.I. (eds) [https://link.springer.com/referencework/10.1007/978-1-4899-7687-1 "Encyclopedia of Machine Learning and Data Mining]. Springer, Boston, MA | ||
** QUOTE: [[Explanation-based learning (EBL)]] involves using [[prior knowledge]] to explain (“prove”) why the [[training example]] has the [[label]] it is given and using this explanation to guide the [[learning]]. Since the explanations are often able to pinpoint the [[feature]]s of the example that justify its [[label]], [[EBL]] techniques are able to get by with much fewer number of [[training examples]]. On the flip side, unlike general [[classification learner]]s, [[EBL]] requires [[prior knowledge]] (aka “domain theory/model”) in [[addition]] to [[labeled training example]]s – a requirement that is not easily met in some scenarios. Since many [[planning]] and [[problem-solving agent]]s do start with [[declarative domain theories]] (consisting at least of descriptions of actions along with their preconditions and effects), [[EBL]] has been a popular [[learning technique]] for [[planning]]. | ** QUOTE: [[Explanation-based learning (EBL)]] involves using [[prior knowledge]] to explain (“prove”) why the [[training example]] has the [[label]] it is given and using this explanation to guide the [[learning]]. Since the explanations are often able to pinpoint the [[feature]]s of the example that justify its [[label]], [[EBL]] techniques are able to get by with much fewer number of [[training examples]]. On the flip side, unlike general [[classification learner]]s, [[EBL]] requires [[prior knowledge]] (aka “domain theory/model”) in [[addition]] to [[labeled training example]]s – a requirement that is not easily met in some scenarios. Since many [[planning]] and [[problem-solving agent]]s do start with [[declarative domain theories]] (consisting at least of descriptions of actions along with their preconditions and effects), [[EBL]] has been a popular [[learning technique]] for [[planning]]. | ||
Revision as of 04:44, 7 February 2018
An Explanation-based Learning for Planning is a planning task based on explanation-based learning.
References
2017
- (Kambhampati & Yoon, 2017) ⇒ Kambhampati S., Yoon S. (2017). "Explanation-Based Learning for Planning". In: Sammut C., Webb G.I. (eds) "Encyclopedia of Machine Learning and Data Mining. Springer, Boston, MA
- QUOTE: Explanation-based learning (EBL) involves using prior knowledge to explain (“prove”) why the training example has the label it is given and using this explanation to guide the learning. Since the explanations are often able to pinpoint the features of the example that justify its label, EBL techniques are able to get by with much fewer number of training examples. On the flip side, unlike general classification learners, EBL requires prior knowledge (aka “domain theory/model”) in addition to labeled training examples – a requirement that is not easily met in some scenarios. Since many planning and problem-solving agents do start with declarative domain theories (consisting at least of descriptions of actions along with their preconditions and effects), EBL has been a popular learning technique for planning.