講 題：Matrix visualization: review, applications and perspectives
主講人：吳漢銘 副教授(國立政治大學統計學系)
時 間：2023年05月18日（星期四）下午02：10 - 04：00
地 點：B302A（淡水校園商管大樓）
茶 會：2023年05月18日（星期四）下午01：30 (商管大樓 B1102)
摘 要
A matrix visualization (MV), also known as a heatmap, is a tool that presents data matrices through colored maps. This visualization technique is particularly useful for investigating the grouping and clustering structure of large-scale, high-dimensional datasets without reducing their dimensions. As an exploratory data analysis (EDA) tool, it has been extensively used in various fields, including bioinformatics, social science, and computer science. In this study, we first review representative works on MV-related techniques and software development. Next, we describe the main procedures of MV within the framework of generalized association plots (GAPs), along with some recent advances. Specifically, we introduce and apply raw data from cellular probes. Additionally, we discuss MV generalizations in the context of visual data mining and highlight some innovative approaches for future research. (This study is based on a joint work with Chun-houh Chen from the Institute of Statistical Science at Academia Sinica.)
Key words: Exploratory Data Analysis, Generalized Association Plots, Heatmap, Symbolic Data Analysis.