Growth and development in early childhood certainly affects how a child is when reaching adulthood both in terms of mental, physical, and intelectual. In the development phase of course not all children experience normal development, there may be a developmental disorder. One developmental disorder that is often experienced in early childhood is ADHD (Attention Deficit Hyperactivity Disorder). For ADHD itself there are three types, among others Inattention, Impulsive, and Hyperactivity. In this research will be identification type of ADHD based on symptoms that appear by using method of classification of Modified K-Nearest Neighbor (MKNN). MKNN method is one method of development of the KNN method, which distinguishes the MKNN there is a validity process and also weight voting of each type to be classified. In this study will be done type identification consisting of 4 types include Inattention, Impulsive, Hyperactivity, and Not ADHD. The results of this study indicate that MKNN method can identify ADHD type well when the data used is 80 data with 20 test data, K = 3 with 90% accuracy. In this study also proves that MKNN method tends to be lower accuracy than KNN method, it is caused by low validity value which will affect weight voting and also accuracy.
Copyrights © 2019