Honors Principles of Data Mining

Catalog description

Real-world datasets are often large and do not fit in main memory, requiring the use of databases. In this honors course, students will go beyond the basic principles and algorithms of data mining, which often assume datasets fit in main memory. Students will engage in creative technical problem solving by adapting these basic principles and algorithms to mine real-world datasets that are stored in different types of databases (relational, document-oriented, graph-oriented). Students will also investigate and reflect on the legal and ethical issues that arise when mining (large) datasets. Topics include the knowledge discovery process; data exploration and cleaning; and data mining algorithms. The emphasis is on how to use data mining to describe datasets rather than making predictions. A term project (presentation and paper) analyzing legal and ethical issues underlying data preparation and mining over large databases is required.