Dimension Reduction Using Simplex Projection and PCA for Multiclass Classification
Hong Zhang · 2025
We propose a dimension reduction method for classification problems with multiple classes by combining principal component analysis and a projection to the simplex of class centers. It is shown that this extended simplex projection method incurs no information loss if the class distributions are Gaussian with a constant covariance matrix. Experimental results using a high dimensional data set showed good performance of the dimension reduction method.