Abstract. Principal Component Analysis (PCA) is one of the most popular techniques for dimensionality reduction of multivariate data points with application areas covering many bra...
We present a tutorial survey on some recent approaches to unsupervised machine learning in the context of statistical pattern recognition. In statistical PR, there are two classica...
Most current anomaly Intrusion Detection Systems (IDSs) detect computer network behavior as normal or abnormal but cannot identify the type of attacks. Moreover, most current intr...
Abstract. In this paper we introduce a new error measure, integrated reconstruction error (IRE) and show that the minimization of IRE leads to principal eigenvectors (without rotat...
This paper presents an algebro-geometric solution to the problem of segmenting an unknown number of subspaces of unknown and varying dimensions from sample data points. We represen...