Selecting Quality Broiler Chicken using Data Mining Technique

Abstract

One of the keys of the Broiler farm is in the initial process of choosing the chicken seeds or known as the day DoC. Some criteria are free special diseases of the mushroom, derived from good quality chicken and so on. The problem is that many electoral processes can make it difficult for farmers and companies to choose good quality broiler chickens. Plus the limited ability of farmers to choose broiler chickens, because they still choose them manually. The research is divided into 3 phases i.e. using AHP to sort the criteria of weights that exist among them the first order free of disease with a weight of 0.207, normal body with a weight of 0.174, weight according to the standard with weights 0.172 and Beyond. While Fuzzy CMeans is used for cluster processes in monitoring the quality level of each chicken nursery process by dividing 3 clusters of cluster 1 as much as 410 breeders (96.70%), a cluster of 2 as many as 12 breeders (2.83%) and cluster 3 as many as 2 farmers (0.47%). The second phase is used by the Decision Tree C 4.5 to process the classification of the selection process whether it enters the category A (good), B (Enough) and C (less). While the third phase is still in development research. The result of this classification process gives the info that the Decision Tree C 4.5 works well marked with a validation accuracy value of about 92.03% with a RoC value of about 0881. This research provides a solution for companies to be able to monitor from 3 parts, they are the inputs, processes and outputs of the farmer's chicken nursery. Contributions for the company are able to reduce losses caused by the fault of the three sections so that the company's profits will increase.