There are many different proposed attribute ranking and selection methods. Hall and Holmes et. al analyzed numerous of these methods and recognized the ones that achieved finest results [] as “Correlation-Based Feature Selection” [] “Information Gain Correlation”, “Wrapper Subset Evaluation”[] , “Recursive Elimination of Features” [] , “Consistency-Based Subset Evaluation” [] .The outcomes help the researchers to select the most appropriate method for attribute selection. Some of the methods are time consuming and their performance is highly dependent on the characteristics of the data. Therefore there isn’t any method that outperforms the others and analysis need to be made for each specific learning problem [] . The …show more content…
Definition 1.1 A random forest is a classifier consisting of a collection of tree structured classifiers { , k=1, ...} where the are independent identically distributed random vectors and each tree casts a unit vote for the most popular class a input x .
Use of the Strong Law of Large Numbers shows that they always converge so that overfitting is not a problem [] . The accuracy of a random forest depends on the strength of the individual tree classifiers and a measure of the dependence between