2007-11-02
org.kosen.entty.User@1c49dfb0
전태균(vuscon)
With the advent of the internet and rapid development of information technologies, we frequently encounter the flood of (high-dimensional) data in many disciplines such as images, biological data, and retail customer information, and so forth. To handle such high-dimensional data efficiently, which is very difficult to build a classifier with good generalization power especially when the number of training data is small, we are developing new dimension reduction and sparse representation methods. We are also trying to understand more accurately the decision process of living things in order to design robust algorithms in learning, optimization and parallel processing, which will be improved versions of genetic algorithms, particle swarm optimization, and neural networks. Image processing with level set methods and wavelets is another research area of our group. By combining well the machine learning, the image processing and the optimization techniques, we are building robust decision support systems in the biomedical, the ubiquitous computing, and environmental engineering fields.