Discovering Interesting Rules from Financial Data
Przemysław Sołdacki, Grzegorz Protaziuk · 2002
In this paper problem of mining data with weights and finding association rules is presented. Some applications are discussed, especially focused on financial data. Solutions of the problem are analyzed. A few approaches are proposed and compared. Pruning based on measures of rules interestingness is described and some measures proposed in literature are shown. Influence of data weights on these measures is also discussed.