Results 1 
2 of
2
Feature Selection via Mathematical Programming
, 1997
"... The problem of discriminating between two finite point sets in ndimensional feature space by a separating plane that utilizes as few of the features as possible, is formulated as a mathematical program with a parametric objective function and linear constraints. The step function that appears in th ..."
Abstract

Cited by 59 (22 self)
 Add to MetaCart
The problem of discriminating between two finite point sets in ndimensional feature space by a separating plane that utilizes as few of the features as possible, is formulated as a mathematical program with a parametric objective function and linear constraints. The step function that appears in the objective function can be approximated by a sigmoid or by a concave exponential on the nonnegative real line, or it can be treated exactly by considering the equivalent linear program with equilibrium constraints (LPEC). Computational tests of these three approaches on publicly available realworld databases have been carried out and compared with an adaptation of the optimal brain damage (OBD) method for reducing neural network complexity. One feature selection algorithm via concave minimization (FSV) reduced crossvalidation error on a cancer prognosis database by 35.4% while reducing problem features from 32 to 4. Feature selection is an important problem in machine learning [18, 15, 1...
Efficient Adaptive Learning For Classification Tasks With Binary Units
 NEURAL COMPUTATION
, 1998
"... A new incremental learning algorithm for classification tasks, called NetLines, well adapted for both binary and realvalued input patterns is presented. It generates small compact feedforward neural networks with one hidden layer of binary units and binary output units. A convergence theorem ensure ..."
Abstract
 Add to MetaCart
A new incremental learning algorithm for classification tasks, called NetLines, well adapted for both binary and realvalued input patterns is presented. It generates small compact feedforward neural networks with one hidden layer of binary units and binary output units. A convergence theorem ensures that solutions with a finite number of hidden units exist for both binary and realvalued input patterns. An implementation for problems with more than two classes, valid for any binary classifier, is proposed. The generalization error and the size of the resulting networks are compared to the best published results on wellknown classification benchmarks. Early stopping is shown to decrease overfitting, without improving the generalization performance.