Abstract
In this paper we present a clinically modified info-fuzzy network (IFN) algorithm and a modified composite association rule algorithm for the analysis of discrete valued clinical data obtained at the Westmead Fertility Clinic. The clinically modified IFN (CMIFN) algorithm takes into account the clinical significance of an attribute in relation to a specified target attribute as well as its statistical significance. This gives us the flexibility for exploring the data set according to the clinical question and hypothesis. The MCAR algorithm gives the flexibility of selecting the composition of the “if” node. It also recursively incorporates all relevant attributes. The results show that the MCAR algorithm is marginally better. On the other hand we note that the CMIFN has the ability to produce negative associations as well as positive and nil associations.
Recommended Citation
Davis, Joseph; Illingworth, Peter; and Salam, A., "Applications of data mining techniques in assisted reproductive technology" (2005). ACIS 2005 Proceedings. 16.
https://aisel.aisnet.org/acis2005/16