Data Visualization and Clustering: an application to gene expression data

Abstract In this work we present a multi-step process that, starting from raw datasets, brings them through preprocessing, preclustering and agglomerative clustering stages, exploiting a visual and interactive environment for data analysis and exploration. At the core of the process lies the idea of subdividing the process of data clusterization into two main steps: the first one aimed to reduce the size of data and the second one to present the reduced and humanunderstandable dataset to the user. This last step allows him to participate in the process of clustering and helps him figuring out the data underlying structures. The approach is actually implemented in a group of user-friendly tools under the MATLAB environment, featuring a number of classical and novel data processing, visualization, assesment and interaction methods.