Biomimetic Pattern Recognition in Cancer Detection
Author:
leonila lagunesMentor:
Charles Lee, Professor of Mathematics, California State University FullertonBiomimetic Pattern Recognition (BPR) is a technique that creates a hyper-dimensional geometric body by mimicking a biological system and uses it for classification. BPR is based on the Principle of Homology-Continuity, which assumes members of a class are connected by means of gradual evolution. Particularly, elements of the same type can be topologically assembled as nodes and the resulting skeleton-like structure is covered with a tissue layer to form a complex biological organism. The end product can subsequently be used for pattern recognition and classification. In recent years, BPR has been used in voice recognition, facial recognition, and image restoration. In our research, we develop several BPR algorithms and apply them to DNA microarray data to detect different types of cancers.