Prof. Kwong Chung-ping
Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong
Data, small size or huge volume, are everywhere. A personal work schedule belongs to the former, the entire networked webpages is a good example of the latter. Traditionally we used geometry and algebra to analyze and> transform data. For instance, PCA (Principal Component Analysis) has been widely applied to estimate the volatility in the Black-Scholes option pricing equation, and the well-known Google search engine is essentially an eigenvector algorithm. On the other hand, the role of topology in handling data has been paltry until recently. In this talk we will introduce the basic principle of topological methods for data analysis. Three specific areas of applications will be discussed, namely, information retrieval, concept analysis, and dimension reduction.