Corpus ID: 7255690

Bridging the Gap between Distance and Generalisation

@inproceedings{Estruch2008BridgingTG,
  title={Bridging the Gap between Distance and Generalisation},
  author={V. Estruch and C. Ferri and J. Hernandez},
  year={2008}
}
Distance-based and generalisation-based methods are two families of artificial intelligence techniques that have been successfully used over a wide range of real-world problems. In the first case, general algorithms can be applied to any data representation by just changing the distance. The metric space sets the search and learning space, which is generally instance-oriented. In the second case, models can be obtained for a given pattern language, which can be comprehensible. The generality… Expand
Towards the definition of learning systems with configurable operators and heuristics ?
The number and performance of machine learning techniques dealing with complex, structured data has considerably increased in the past decades. However, the performance of these systems is usuallyExpand
On the definition of a general learning system with user-defined operators
TLDR
This paper proposes a learning paradigm where users can write (or adapt) their operators, according to the problem, data representation and the way the information should be navigated, which makes it possible to have a more flexible mapping between old and new problems. Expand
Learning with Configurable Operators and RL-Based Heuristics
TLDR
The architecture can be seen as a system for writing machine learning systems' or to explore new operators, where users can write (or adapt) their operators, according to the problem, data representation and the way the information should be navigated. Expand
Incremental and developmental perspectives for general-purpose learning systems
TLDR
This thesis contributes with a general-purpose learning system that meets several desirable characteristics in terms of expressiveness, comprehensibility and versatility, and analyses whether the use of intelligence tests for AI evaluation is a much better alternative to most task-oriented evaluation approaches in AI. Expand
Bridging the gap between distance and generalisation: symbolic learning in metric spaces
El aprendizaje automatico es la disciplina que se centra en el diseno de algoritmos que pueden aprender a partir de datos recogidos de la experiencia. Muchos de estos algoritmos son fundamentales enExpand

References

SHOWING 1-10 OF 51 REFERENCES
Distance Based Generalisation
TLDR
This paper defines several conditions which, in its view, a sensible distance-based generalisation must have, and shows that there are pattern languages and generalisation operators which comply with these properties for typical data types: nominal, numerical, sets and lists. Expand
Distance-Based Generalisation Operators for Graphs
Distances and similarity functions between structured datatypes, such as graphs, have been widely employed in machine learning since they are able to identify similar cases or prototypes from whichExpand
Generalisation Operators for Lists Embedded in a Metric Space
TLDR
This paper investigates the way in which, given a pattern language, a pair of generalisation operator and distance which are consistent are defined, and defines the notion of (minimal) distance-based generalisation operators for lists. Expand
An Instantiation of Hierarchical Distance-Based Conceptual Clustering for Propositional Learning
TLDR
This work proposes a set of generalisation operators and distances for the data types mentioned before and analyses the properties by them satisfied on the basis of three different levels of agreement between the clustering hierarchy obtained from the linkage distance and the hierarchy obtained by using generalisation Operators. Expand
Unifying instance-based and rule-based induction
TLDR
In an extensive empirical study, RISE consistently achieves higher accuracies than state-of-the-art representatives of both its parent approaches, as well as a decision tree learner (C4.5). Expand
A Taxonomy of Similarity Mechanisms for Case-Based Reasoning
  • P. Cunningham
  • Computer Science
  • IEEE Transactions on Knowledge and Data Engineering
  • 2009
TLDR
This paper presents a taxonomy that organizes new similarity mechanisms and more established similarity mechanisms in a coherent framework and presents research on kernel-based learning is a rich source of novel similarity representations because of the emphasis on encoding domain knowledge in the kernel function. Expand
A Framework for Defining Distances Between First-Order Logic Objects
TLDR
This paper develops a general framework for distances between distances between objects and reports a preliminary evaluation. Expand
Distance measures for point sets and their computation
TLDR
The metric infimum of the minimum link measure is presented, a new distance function which is more appealing than the other distance functions mentioned and shown to be in NP for a broad class of instances; it is NP-hard for a natural problem class. Expand
Deriving distance metrics from generality relations
TLDR
This work proves a theorem that shows how a distance metric can be derived from a partial order (and a corresponding size on patterns) under mild conditions and shows how well-known distance metrics for sets, strings, trees and graphs can bederived from their generality relation. Expand
A nearest hyperrectangle learning method
This paper presents a theory of learning called nested generalized exemplar (NGE) theory, in which learning is accomplished by storing objects in Euclidean n-space, En, as hyperrectangles. TheExpand
...
1
2
3
4
5
...