Initial state prediction in planning

Senka  Krivic; Michael  Cashmore; Bram Ridder; Daniele  Magazzeni; Sandor Szedmak; Justus  Piater

Initial state prediction in planning

Senka Krivic, Michael Cashmore, Bram Ridder, Daniele Magazzeni, Sandor Szedmak, Justus Piater

Computer And Information Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

9 Downloads (Pure)

Abstract

While recent advances in offline reasoning techniques and online execution strategies have made planning under uncertainty more robust, the application of plans in partially-known environments is still a difficult and important topic. In this paper we present an approach for predicting new information about a partially-known initial state, represented as a multi- graph utilizing Maximum-Margin Multi-Valued Regression. We evaluate this approach in four different domains, demonstrating high recall and accuracy.

Original language	English
Title of host publication	The AAAI-17 Workshop on Knowledge-Based Techniques for Problem Solving and Reasoning - Technical Report
Place of Publication	Menlo Park, US-CA.
Number of pages	8
Volume	WS-17-12
Publication status	Published - 5 Feb 2017

Keywords

planning
initial state prediction

Access to Document

Krivic-etal-AAAI2017-Initial-state-prediction-planningAccepted author manuscript, 1.38 MB

Cite this

@inproceedings{91882f8a075c4d7c8eafb99acbb3fc4a,

title = "Initial state prediction in planning",

abstract = "While recent advances in offline reasoning techniques and online execution strategies have made planning under uncertainty more robust, the application of plans in partially-known environments is still a difficult and important topic. In this paper we present an approach for predicting new information about a partially-known initial state, represented as a multi- graph utilizing Maximum-Margin Multi-Valued Regression. We evaluate this approach in four different domains, demonstrating high recall and accuracy.",

keywords = "planning, initial state prediction",

author = "Senka Krivic and Michael Cashmore and Bram Ridder and Daniele Magazzeni and Sandor Szedmak and Justus Piater",

year = "2017",

month = feb,

day = "5",

language = "English",

isbn = "9781577357865",

volume = "WS-17-12",

booktitle = "The AAAI-17 Workshop on Knowledge-Based Techniques for Problem Solving and Reasoning - Technical Report",

}

TY - GEN

T1 - Initial state prediction in planning

AU - Krivic, Senka

AU - Cashmore, Michael

AU - Ridder, Bram

AU - Magazzeni, Daniele

AU - Szedmak, Sandor

AU - Piater, Justus

PY - 2017/2/5

Y1 - 2017/2/5

N2 - While recent advances in offline reasoning techniques and online execution strategies have made planning under uncertainty more robust, the application of plans in partially-known environments is still a difficult and important topic. In this paper we present an approach for predicting new information about a partially-known initial state, represented as a multi- graph utilizing Maximum-Margin Multi-Valued Regression. We evaluate this approach in four different domains, demonstrating high recall and accuracy.

AB - While recent advances in offline reasoning techniques and online execution strategies have made planning under uncertainty more robust, the application of plans in partially-known environments is still a difficult and important topic. In this paper we present an approach for predicting new information about a partially-known initial state, represented as a multi- graph utilizing Maximum-Margin Multi-Valued Regression. We evaluate this approach in four different domains, demonstrating high recall and accuracy.

KW - planning

KW - initial state prediction

M3 - Conference contribution book

SN - 9781577357865

VL - WS-17-12

BT - The AAAI-17 Workshop on Knowledge-Based Techniques for Problem Solving and Reasoning - Technical Report

CY - Menlo Park, US-CA.

ER -

Initial state prediction in planning

Abstract

Keywords

Access to Document

Fingerprint

Cite this