Biography: BaoXiang Liu is a professor at North China university of science and technology. Since 2004 he is dean of College of Science. Since 2010 he is director of the institute of statistics in Hebei province. His research interests include Rough sets theory and Knowledge discovery application, machine learning, and Fuzzy decision. He is now director of key laboratory of Hebei province, head of the key subject of applied mathematics in Hebei province, and expert of the national natural fund evaluation.
Speech Title: Data Mining Based on Interval Concept Lattice
Abstract: The report is part of the research for the national natural fund “theory of interval concept lattice and application of dynamic optimization for rough control rules” .Interval concept lattice is the development of a classic concept lattice and rough concept lattice, it is designed to solve the contradiction between mining cost, efficiency and reliability of rule mining. Interval concept lattice is the set of objects of concept extension within the interval [α，β](0≤α≤β≤1) which meet connotation attributes, and it is stronger than the rough concept lattice's ability to deal with uncertain knowledge. Mainly introduces the structure characteristics and the corresponding properties of interval concept lattice and the theorem, according to the characteristics of interval concept lattice, how to build the attribute reduction model and efficient algorithm based on association rules, and put forward a optimization model and parameter range of interval [α, β] parameters under control of classification rules.