jcolibri.method.retrieve.selection.diversity
Class BoundedGreedySelection

java.lang.Object
  extended by jcolibri.method.retrieve.selection.diversity.BoundedGreedySelection

public class BoundedGreedySelection
extends java.lang.Object

Tries to reduce the complexity of the greedy selection algorithm first selecting the best b·k cases according to their similarity to the query and then applies the greedy selection method to these cases.

See:

B. Smyth and P. McClave. Similarity vs. diversity. In ICCBR '01: Proceedings of the 4th International Conference on Case-Based Reasoning, pages 347-361, London, UK, 2001. Springer-Verlag.

Version:
1.0
Author:
Juan A. Recio-Garcia, Developed at University College Cork (Ireland) in collaboration with Derek Bridge.

Constructor Summary
BoundedGreedySelection()
           
 
Method Summary
static java.util.Collection<CBRCase> boundedGreddySelection(java.util.Collection<RetrievalResult> cases, CBRQuery query, NNConfig simConfig, int k, int bound)
          Executes the algorithm
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

BoundedGreedySelection

public BoundedGreedySelection()
Method Detail

boundedGreddySelection

public static java.util.Collection<CBRCase> boundedGreddySelection(java.util.Collection<RetrievalResult> cases,
                                                                   CBRQuery query,
                                                                   NNConfig simConfig,
                                                                   int k,
                                                                   int bound)
Executes the algorithm

Parameters:
cases - to retrieve from
query - to compare
simConfig - is the knn config that defines the k returned cases
bound - to create the bounded retrieval set
k - is the number of cases to select
Returns:
k cases

GAIA - Group for Artificial Intelligence Applications
http://gaia.fdi.ucm.es