Application of Intuitionitic Fuzzy C-means Clustering Method in Cocoa Beans Data

• Main Author: Dani, Yasi
• Format: Theses
• Language: Indonesia
• Subjects: Matematika
• Online Access: https://digilib.itb.ac.id/gdl/view/33754
• Institution: Institut Teknologi Bandung

Clustering is the process of grouping data into clusters based on certain criteria so that the objects in a cluster have a high degree of similarity to each other. Intuitionistic fuzzy C-means (IFCM) is included in one of the clustering technique, where the existence of each data point in a cluster is determined by membership and non membership with hesitancy degree. IFCM clustering process is similar to the Fuzzy C-means clustering (FCM) both algorithms are based on minimization of the objective function that describe the distance between cluster center and data points weighted by the degrees of membership. However, to incorporate intuitionistic fuzzy property we modify the membership degree of conventional fuzzy by including the hesitation degree. In this thesis, the IFCM is used to classify the cocoa beans data arising from six treatments: no fermented and roasted, fermented in the eld and no roasted, fermented in laboratory and no roasted, no fermented and roasted, fermented in eld and roasted, and fermented in the laboratory and roasted. First step of our work is to reduce the experimental data from three data sets to become one data set, for each treatment. this reducing process will be worked out by two proposed methods: reduction of the outliers, and direct fuzzy clustering. Next the IFCM is applied to these reduced data. The goodness of the obtained clusters will be measured by the Xie-Beni index.