-
Jacky Baltes and Bruce MacDonald.
An Integrated Planning Representation using Macros, Abstractions, and Cases.
In Michael R. Lowry, editor,
Proceedings of the Workshop on Change of Representation and Problem Reformulation,
Moffet Field, CA 94025,USA,
pages 1-10,
April 1992.
NASA Ames Research Center,
NASA Ames Research Center.
Abstract: |
Planning will be an essential part of future autonomous robots and integrated intelligent systems. After giving a brief introduction to the classical planning paradigm, this paper focuses on learning problem solving knowledge in planning systems. A general weak method for learning useful operators is the creation of macros. The paper first describes a novel approach to the selection and dynamic filtering of macros. The dynamic filtering approach is suggested for controlling the creation of operators. A new planning representation is proposed that uses a common representation for macros, abstractions, and cases. A general operator is represented by sequences of primitive or non--primitive operators. A macro is equivalent to a sequence of primitive, executable, operators with uninstantiated arguments. A case consists of primitive operators with instantiated arguments. An abstract plan is equivalent to a sequence of non--primitive operators at a lower level of abstraction. A learned indexing mechanism allows rapid access to relevant operators. The system is able to use both classical and case--based techniques. The general operators in a successful plan derivation would be assessed for the potential usefulness, and some stored. |
@inproceedings{baltes-1992d,
author = {Jacky Baltes and Bruce MacDonald},
title = {An Integrated Planning Representation using Macros, Abstractions, and Cases},
booktitle = {Proceedings of the Workshop on Change of Representation and Problem Reformulation},
year = 1992,
month = apr,
editor = {Michael R. Lowry},
publisher = {NASA Ames Research Center},
address = {Moffet Field, CA 94025,USA},
pages = {1-10},
organization = {NASA Ames Research Center},
abstract = {Planning will be an essential part of future autonomous robots and integrated intelligent systems. After giving a brief introduction to the classical planning paradigm, this paper focuses on learning problem solving knowledge in planning systems. A general weak method for learning useful operators is the creation of macros. The paper first describes a novel approach to the selection and dynamic filtering of macros. The dynamic filtering approach is suggested for controlling the creation of operators. A new planning representation is proposed that uses a common representation for macros, abstractions, and cases. A general operator is represented by sequences of primitive or non--primitive operators. A macro is equivalent to a sequence of primitive, executable, operators with uninstantiated arguments. A case consists of primitive operators with instantiated arguments. An abstract plan is equivalent to a sequence of non--primitive operators at a lower level of abstraction. A learned indexing mechanism allows rapid access to relevant operators. The system is able to use both classical and case--based techniques. The general operators in a successful plan derivation would be assessed for the potential usefulness, and some stored.},
pdf = {http://aalab.cs.umanitoba.ca/%7ejacky/Publications/pdf/baltes-1992d.pdf}
}
-
Jacky Baltes and Bruce MacDonald.
Case--based Meta Learning: Sustained Learning supported by a Dynamically Biased Version Space.
In Diana Gordon, editor,
Proceedings of the ML 92 Workshop on Biases in Inductive Learning,
July 1992.
Abstract: |
It is well--recognized that in practical inductive learning systems the search for a concept must be heavily biased. In addition the bias must be dynamic, adapting to the current learning problem. Another important requirement is sustained learning, allowing transfer from known tasks to new ones. Previous work on dynamic bias has not explicitly addressed learning transfer, while previous case--based learning research suffers from a variety of problems. This paper presents a method of Case--Based Meta Learning (CBML), in which the cases are concepts, rather than instances, and retrieved similar concepts are used as a skeletal version space to speed up learning. CBML is independent of the concept representation language. The CBML--Clerk system, which learns repetitive operating system tasks, is presented as a demonstration. |
@inproceedings{baltes-1992b,
key = {ML 92 workshop},
author = {Jacky Baltes and Bruce MacDonald},
title = {Case--based Meta Learning: Sustained Learning supported by a Dynamically Biased Version Space},
booktitle = {Proceedings of the ML 92 Workshop on Biases in Inductive Learning},
year = 1992,
month = {July},
editor = {Diana Gordon},
abstract = {It is well--recognized that in practical inductive learning systems the search for a concept must be heavily biased. In addition the bias must be dynamic, adapting to the current learning problem. Another important requirement is sustained learning, allowing transfer from known tasks to new ones. Previous work on dynamic bias has not explicitly addressed learning transfer, while previous case--based learning research suffers from a variety of problems. This paper presents a method of Case--Based Meta Learning (CBML), in which the cases are concepts, rather than instances, and retrieved similar concepts are used as a skeletal version space to speed up learning. CBML is independent of the concept representation language. The CBML--Clerk system, which learns repetitive operating system tasks, is presented as a demonstration.},
pdf = {http://aalab.cs.umanitoba.ca/%7ejacky/Publications/pdf/baltes-1992b.pdf}
}
-
Jacky Baltes and Bruce MacDonald.
Case--based Meta Learning: Using a Dynamically Version Space in Sustained Learning.
In Janice Glasgow and Robert Hadley, editors,
Proceedings Ninth Canadian Conference on Artificial Intelligence,
Palo Alto, California,
pages 228-235,
May 1992.
Canadian Society for Computational Studies of Intelligence,
Morgan Kaufman Publishers Inc..
Abstract: |
It is well--recognized that in practical inductive learning systems the search for a concept must be heavily biased. In addition the bias must be dynamic, adapting to the current learning problem. Another important requirement is sustained learning, allowing transfer from known tasks to new ones. Previous work on dynamic bias has not explicitly addressed learning transfer, while previous case--based learning research suffers from a variety of problems. This paper presents a method of Case--Based Meta Learning (CBML), in which the cases are concepts, rather than instances, and retrieved similar concepts are used as a skeletal version space to speed up learning. CBML is independent of the concept representation language. The CBML--Clerk system, which learns repetitive operating system tasks, is presented as a demonstration. |
@inproceedings{baltes-1992,
author = {Jacky Baltes and Bruce MacDonald},
title = {Case--based Meta Learning: Using a Dynamically Version Space in Sustained Learning},
booktitle = {Proceedings Ninth Canadian Conference on Artificial Intelligence},
organization = {Canadian Society for Computational Studies of Intelligence},
year = {1992},
month = may,
editor = {Janice Glasgow and Robert Hadley},
publisher = {Morgan Kaufman Publishers Inc.},
address = {Palo Alto, California},
pages = {228-235},
abstract = {It is well--recognized that in practical inductive learning systems the search for a concept must be heavily biased. In addition the bias must be dynamic, adapting to the current learning problem. Another important requirement is sustained learning, allowing transfer from known tasks to new ones. Previous work on dynamic bias has not explicitly addressed learning transfer, while previous case--based learning research suffers from a variety of problems. This paper presents a method of Case--Based Meta Learning (CBML), in which the cases are concepts, rather than instances, and retrieved similar concepts are used as a skeletal version space to speed up learning. CBML is independent of the concept representation language. The CBML--Clerk system, which learns repetitive operating system tasks, is presented as a demonstration.},
pdf = {http://aalab.cs.umanitoba.ca/%7ejacky/Publications/pdf/baltes-1992.pdf}
}
-
Ken Barker,
Mark Evans,
and John Anderson.
Quantification of Autonomy in Multi-Agent Systems.
In Proceedings of the AAAI Workshop on Cooperation Among Heterogeneous Agents,
San Jose, CA,
pages 6,
July 1992.
@inproceedings{withbarker,
author = {Ken Barker and Mark Evans and John Anderson},
title = {Quantification of Autonomy in Multi-Agent Systems},
booktitle = {Proceedings of the AAAI Workshop on Cooperation Among Heterogeneous Agents},
year = {1992},
pages = {6},
address = {San Jose, CA},
month = {July}
}
-
Bruce MacDonald and Jacky Baltes.
Research in instructable Systems.
In Machine Learning Workshop at AI/GI/VI '92,
May 1992.
@inproceedings{baltes-1992e,
key = {ML workshop at CAI 92},
author = {Bruce MacDonald and Jacky Baltes},
title = {Research in instructable Systems},
booktitle = {Machine Learning Workshop at AI/GI/VI '92},
year = {1992},
month = {May}
}