Association Rules Analysis for Continuous Chicken Egg Traits Dataset

Abstract

This study aims to apply the Apriori association rule algorithm on 14 continuous egg quality traits recorded from 4320 eggs of three commercial white-laying chicken lines. In the study all the continuous data were discretized using Equal-Width-Interval method based the number of intervals obtained with Rice formula. Association rules analysis on the discretized dataset resulted with a total of 349 rules consists of 3 and 4 items. According to the top five rules by support and confidence, some important associations were obtained between the certain value ranges of the traits egg weight, egg width, egg length, shell thickness, and shell breaking strength when compared to the others. The appropriate biological and economic interpretations of the obtained rules may contribute to the poultry industry in practice.