Data Mining

Description

Today it is possible to collect and store vast quantities of data. These data often contain value information and insights. However, it may take human analysists weeks or months to discover the information if they are able to do it at all. Furthermore, so much data exist that most of it is never even analyzed. The goal of data mining is to fill this void by automatically identify models and patterns from these databases that are (1) valid, that is, they hold on new data with some certainty, (2) novel, that is, they are non-obvious, (3) useful, that is, they are actionable, and (4) understandable. that is humans can interpret them. In order to do this, data mining, also called knowledge discovery in databases (KDD), combines ideas from the fields of machine learning, databases, statistics, visualization, and many other fields. The goal of this course is to provide a broad survey of several important and well-know fields of data mining and to develop an overall sense of how to extract information from data in a systematic way. It tries to give inisght into the challenges faced by data miners and the inner workers of specific data mining algorithms as well as provide some understanding about why data mining is important and interesting. The course consists of lectures, readings and exercises sessions. The exercise sessions reinforce the central concepts covered during class and give students some experience working with publicly available data mining tools. The course requires knowledge of machine learning.