This paper proposes and explains a data treatment technique to improve the accuracy of a neural network estimator in regression problems, where multi-dimensional input data set is...
— Local linearizations are ubiquitous in the control of robotic systems. Analytical methods, if available, can be used to obtain the linearization, but in complex robotics system...
The problem of learning a sparse conic combination of kernel functions or kernel matrices for classification or regression can be achieved via the regularization by a block 1-norm...
Francis R. Bach, Romain Thibaux, Michael I. Jordan
Most of the current Intrusion Detection Systems (IDS) examine all data features to detect intrusion or misuse patterns. Some of the features may be redundant or contribute little (...
Srilatha Chebrolu, Ajith Abraham, Johnson P. Thoma...
The usual way to compare testing strategies, whether theoretically or empirically, is to compare the number of faults they detect. To ascertain definitely that a testing strategy...
Ilinca Ciupa, Bertrand Meyer, Manuel Oriol, Alexan...