In multiple instance learning (MIL), how the instances determine the bag-labels is an essential issue, both algorithmically and intrinsically. In this paper, we show that the mech...
In this paper, I propose a genetic algorithm (GA) approach to instance selection in artificial neural networks (ANNs) for financial data mining. ANN has preeminent learning abilit...
The purpose of this paper is to provide a preliminary report on the rst broad-based experimental comparison of modern heuristics for the asymmetric traveling salesmen problem ATSP....
Jill Cirasella, David S. Johnson, Lyle A. McGeoch,...
Multi-instance learning deals with problems that treat bags of instances as training examples. In single-instance learning problems, dimensionality reduction is an essential step ...
: Countably based filter spaces have been suggested in the 1970's as a model for recursion theory on higher types. Weak limit spaces with a countable base are known to be the ...