Proactive Data Mining: A General Approach and Algorithmic Framework

Abstract

In the previous section we presented several important data mining concepts. In this chapter, we argue that with many state-of-the-art methods in data mining, the overly-complex responsibility of deciding on this action or that is left to the human operator. We suggest a new data mining task, proactive data mining. This approach is based on supervised learning, but focuses on actions and optimization, rather than on extracting accurate patterns. We present an algorithmic framework for tackling the new task. We begin this chapter by describing our notation.