Using Forward / Backword Decision Trees on GSM441161 and GSE32863 Microarray Data

Microarray Experiments (MEs) are done on solid surface called probes, each array contains thousands of genes. ME measures the transcription of nucleic acid, which is a macromolecules storing information about protein production, protein is the main components of all human tissues, which means that any changes in ME will cause dysfunctions of the concerned studied part of the organ, since in a ME we can study only a unique characteristic. Microarray data are represented with few samples and huge amount of genes. In this work, we used embedded method with random forest in two different ways, backward and forward with microarray data, for the purpose of validation we used the accuracy.