Tree models are valuable tools for predictive modeling and data mining. Traditional tree-growing methodologies such as CART are known to suffer from problems including greediness,...
Algebraic randomization techniques can be applied to hybrid symbolic-numeric algorithms. Here we consider the problem of interpolating a sparse rational function from noisy values...
Database integration of data mining has gained popularity and its significance is well recognized. However, the performance of SQL based data mining is known to fall behind specia...