Bayesian text classifiers face a common issue which is referred to as data sparsity problem, especially when the size of training data is very small. The frequently used Laplacian...
Most existing decision tree inducers are very fast due to their greedy approach. In many real-life applications, however, we are willing to allocate more time to get better decisi...
The majority of the existing algorithms for learning decision trees are greedy--a tree is induced top-down, making locally optimal decisions at each node. In most cases, however, ...
In multimodal function optimization, niching techniques create diversification within the population, thus encouraging heterogeneous convergence. The key to the effective diversif...
WASA is an intelligent self-assessment agent of WADIES, a web-based learning environment which adapts its material to reflect the needs of the individual learner, whether it is fo...