Description: FASBIR(Filtered Attribute Subspace based Bagging with Injected Randomness) is a variant of Bagging algorithm, whose purpose is to improve accuracy of local learners, such as kNN, through multi-model perturbing ensemble.
Reference: Z.-H. Zhou and Y. Yu. Ensembling local learners through multimodal perturbation. IEEE Transactions on Systems, Man, and Cybernetics - Part B: Cybernetics, 2005, vol.35, no.4, pp.725-735.
A new blind adaptive multiuser detection scheme based on a hybrid of Kalman filter and
subspace estimation is proposed. It is shown that the detector can be expressed as an anchored
signal in the signal subspace and the coefficients can be estimated by the Kalman filter using only
the signature waveform and the timing of the desired user.
％ COMPDIR Computes a search direction in a subspace defined by Z.
％ Helper function for NLCONST.
％ Returns Newton direction if possible.
％ Returns random direction if gradient is small.
％ Otherwise, returns steepest descent direction.
％ If the steepest descent direction is small it computes a negative
％ curvature direction based on the most negative eigenvalue.
％ For singular matrices, returns steepest descent even if small.
