Research Article
An Efficient Ensemble Learning Method for Gene Microarray Classification
Pseudocode 1
The RotBoost pseudocode.
Input | (i) : a training set, where is an matrix containing the input values and is an | N-dimensional column vector containing the class labels. | (ii) : number of attribute subsets (or : number of input attributes contained in each subset). | (iii) : a base learning. | (iv) : number of iterations for Rotation Forest. | (v) : number of iterations for AdaBoost. | (vi) : a data point to be classified. | Training Phase | For | (1) use the steps similar to those in Rotation Forest to compute the Rotation matrix, say, and let be the | training set for classifier . | (2) Initialize the weight distribution over as . | (3) For | (a) According to distribution perform N extractions randomly from with replacement to compose a new set . | (b) Apply to to train a classifier and then compute the error of as | | (c) If then set . and go to step (a); if , then set to continue | the following iterations. | (d) Choose | (e) Update the distribution over as: | | where is a normalization factor being chosen so that is a probability distribution over . | Endfor | (4) Let | Endfor | Output | (i) The class label for x predicted by the final ensemble as | , | where I(·) is an indicator function. |
|