Use of apriori algorithm for discovering association rules in agricultural data mining

Association rule mining consists of the techniques related to finding frequent patterns, associations, correlations, or causal structures among sets of items in transaction databases. Association rules are the rules which aim to predict the occurrence of a specific item based on the occurrences of the other items in a given set of transactions. It is an important tool for finding frequent patterns and co-occurring associations among a collection of items. In this study, an egg quality traits dataset was used as an experimental dataset. Data preprocessing consisting of missing values imputation and discretization was ran on the dataset because before association rules mining requires complete data with discrete features. Apriori algorithm was used in order to obtain the rules among the features. Totally 349 rules were determined on the examined dataset which contains 15 features for 4320 eggs. In order to determine the important ones, the obtained rules were reviewed and also visually inspected for their support, confidence and lift values. According to the results, the first important rules were determined as V1=1, V3=6 ⇒ V2=2; V2=23, V3=23 ⇒ V1=24; V5=10, V3=12, V14=11 ⇒ V9=5. İn this case the first rule states that the egg weight [47.7, 48.6) and length variables [52.6, 53.0) are in the continuous value range, then the egg width [40.0, 40.2) will also be in the continuous value range