Abstract | ||
---|---|---|
The coverage of a learning algorithm is the number of concepts that can be learned by that algorithm form samples of a given size. This paper asks whether good learning algorithms can be designed by maximizing their coverage. This paper extends a previous upper bound on the coverage of any Boolean concept learning algorithm and describes two algorithms- Multi-Balls and Large-Ball- whose coverage of the ID3 and FRINGE algorithms shows that their coverage is far below this bound. Further analysis of Large- Ball shows that although it learns many concepts, these do not seem to be very interesting concepts. Hence, coverage maximization alone does not appear to yields practically- useful learning algorithms. The paper concludes with a definition of coverage within a bias, which suggests a way that coverage maximization could be applied to strengthen weak preference biases. |
Year | DOI | Venue |
---|---|---|
1992 | 10.1016/B978-1-55860-247-2.50007-3 | ML |
Keywords | Field | DocType |
boolean concept,algorithm form sample,interesting concept,good learning algorithm,coverage maximization,weak preference bias,fringe algorithm,useful learning algorithm | Maximum coverage problem,Computer science,Upper and lower bounds,Concept learning,Artificial intelligence,ID3,Machine learning,Maximization | Conference |
Issue | ISBN | Citations |
1 | 1-5586-247-X | 1 |
PageRank | References | Authors |
1.10 | 7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hussein Almuallim | 1 | 547 | 138.58 |
Thomas G. Dietterich | 2 | 9336 | 1722.57 |