A Weighted based Method for Classification and Dimension Reduction Using IWO Algorithm

This paper investigates the problem of feature selection and classification by scaling feature vectors with weight vectors. The main purpose is to find a weight vector associated to each feature showing the relevant importance of each feature. Weight vectors are optimized using Invasive Weed Optimization (IWO) algorithm. The minor features having minimum weights are then removed to reduce the dimensions of feature space. The results of this approach demonstrated better classification rate compared to GA method.

کلیدواژه ها:

Feature selection ، dimension reduction ، IWO Introduction