Cats that are often act as pets to humans is not spared from diseases attack. Skin disease is a common disease suffered by these mammals, if not handled quickly and accurately then the disease can quickly escalate to interfere cat's activity or can even cause death. Early symptoms of skin diseases are sometimes not so visible and not so disturbing, therefore sometimes the cat evem looks fine so the owner is not so concerned. Very limited knowledge of the owner about skin diseases experienced by cats, as well as the many similarities of the symptoms of various skin diseases that are difficult to be identified by the common people became the main reason for the author to conduct research on the diagnosis of skin diseases in cats using the Modified K-Nearest Neighbor method. The Modified K-Nearest Neighbor Method is used for the classification of new data which class is not known based on the nearest k value. The dataset used in this study consisted of 240 cat skin disease data with 14 parameters and 5 different kind of skin diseases, the output of this system in the form of disease diagnosis. The highest accuracy that was obtained based on various testings is 100% at the value of k = 1. From the results of the accuracy, it can be concluded that the Modified K-Nearest Neighbor method can be implemented into the skin disease diagnosis system in cats.
Copyrights © 2017