EtmlSI O : A Program That Learns New Heuristics and Domain Concepts The Nature of Heuristics III : Program Design and Results

@inproceedings{Lenat2005EtmlSIO,
  title={EtmlSI O : A Program That Learns New Heuristics and Domain Concepts The Nature of Heuristics III : Program Design and Results},
  author={Douglas B. Lenat},
  year={2005}
}
The AM program, an early attempt to mechanize learning by discorery, has recently been expanded and extended to set'eral otl, er task. domains. AM's tdtimate failure apparently was due to its inability to discot'er new, powerftd, domain-sl~'cific heuristics for the rarious new fiehts it .ncot'ered. A t that time, it seemed straight-fonvard to simply add "lleuristics" as one more field in which to let A.~I explore, obsert'e, define, and det'elop. That task--learning new het~ristics by discorery… Expand
ARTIFICIAL INTELLIGENCE 31 Theory Formation by Heuristic Search The Nature of Heuristics II : Background and Examples D o u g l a s
Ahwhine h'arning can be categorized along many dimcnsiotts, an important one o/ ~hich is "degree of human guidance or forethought'. Thi~ cotztintaon stretches from rote learning, throughExpand
Toward Automated Discovery in the Biological Sciences
TLDR
The framework described, called the agenda- and justification-based framework, has several properties that are desirable in semiautonomous discovery systems: It provides a mechanism for estimating the plausibility of tasks, it uses heuristics to propose and perform tasks, and it facilitates the encoding of general discovery strategies and the use of background knowledge. Expand
Learning list concepts through program induction
TLDR
A novel concept learning paradigm called Martha's Magical Machines is introduced that captures complex relationships between concepts and accurately predicts that participants learn some transformations more easily than others and that they learn harder concepts more easily using a bootstrapping curriculum focused on their compositional parts. Expand
Distributed Reeective Architectures Thesis Proposal
The autonomy of a system can be deened as its capability to recover from unforeseen diiculties without any user intervention. This thesis proposal addresses a small part of this problem, namely theExpand
Measuring the Diiculty of Speciic Learning Problems
Existing complexity measures from contemporary learning theory cannot be conveniently applied to speciic learning problems (e.g., training sets). Moreover, they are typically non-generic, i.e., theyExpand
Modelling serendipity in a computational context
TLDR
It is feasible to equip computational systems with the potential for serendipity, and that this could be beneficial in varied computational creativity/AI applications, particularly those designed to operate responsively in real-world contexts. Expand
Quantitative Concept Analysis
TLDR
A task analysis is developed which addresses the combinatorial properties of conceptualization, dividing the space of concept-combination mechanisms into distinct prototypes and assessing the generative properties of each. Expand
A Quantitative Reconstruction of Boden's Creativity Theory
TLDR
In this re-working, Boden’s definition of creativity is redeveloped as a qualitative abstraction relating to a formally defined complexity tradeoff and has the capacity to underwrite fully deterministic, process classifications and to offer clearer theoretical guidance for systems-building work. Expand
Autonomous recovery from hostile code insertion using distributed reflection
TLDR
Implementation in a simple virtual world shows that the system can recover from certain kinds of hostile code attacks that cause it to make wrong decisions in its application domain, even if some of its self-monitoring components are also disabled. Expand
Knowledge-directed Adaptation in Multi-level Agents
TLDR
This paper explores agents implemented in Soar that behave and learn at the deliberate and reflective levels of knowledge and control, and hypothesizes that complete, adaptive agents must be able to learn across all three levels. Expand
...
1
2
...

References

SHOWING 1-10 OF 24 REFERENCES
The Nature of Heuristics
  • D. Lenat
  • Computer Science
  • Artif. Intell.
  • 1982
Abstract Builders of expert rule-based systems attribute the impressive performance of their programs to the corpus of knowledge they embody: a large network of facts to provide breadth of scope, andExpand
RLL-1: A Representation Language Language
TLDR
The components of representation languages in general and of RLL-1 itself, in particular, are encoded declaratively as frames, and by modifying these frames, the user can change the semantics of R LL-1's components, and significantly alter the overall character of the Rll-1 environment. Expand
Heuristic Search for New Microcircuit Structures: An Application of Artificial Intelligence
Eurisko is an AI program that learns by discovery. We are applying Eurisko to the task of inventing new kinds of three- dimensional microelectronic devices that can then be fabricated using recentlyExpand
Theory Formation by Heuristic Search
  • D. Lenat
  • Computer Science
  • Artif. Intell.
  • 1983
TLDR
The accretion model of theory formation is presented, and many examples of its use in producing new discoveries in various fields are given, drawn from runs of a program called eurisko, the successor to am, that embodies the Accretion model and uses a corpus of heuristics to guide its behavior. Expand
SOME PHILOSOPHICAL PROBLEMS FROM THE STANDPOINT OF ARTI CIAL INTELLIGENCE
TLDR
The formalism of this paper represents an advance over McCarthy (1963) and Green (1969) in that it permits proof of the correctness of strategies that contain loops and strategies that involve the acquisition of knowledge; and it is also somewhat more concise. Expand
A Man-Machine Theorem-Proving System
TLDR
A man-machine theorem proving system at the University of Texas (Austin) which has been used to prove a few theorems in general topology, built around a modified form of IMPLY, a natural-deduction-like theorem proving technique. Expand
Building expert systems
  • F. Hayes-Roth
  • Engineering, Computer Science
  • Advanced book program
  • 1983
TLDR
This book provides a broad introduction to the concepts and methods necessary for an understanding of how these systems work. Expand
One-gate-wide CMOS Inverter on laser-recrystallized polysilicon
A CMOS inverter having a single gate for both n and p channel devices has been fabricated using bulk silicon for the p channel device and a laser-recrystallized silicon film for the n channel device.Expand
Theory formation by heuristic search ; the nature of heuristics II : Background an d examples
  • Artificial Intelligence 21(1,
  • 1983
Theory formation by heuristic search; the nature of heuristics II: Background and examples
  • Artificial bztelligence
  • 1983
...
1
2
3
...