<![CDATA[In 2017, there are 1,742,285 cannabis (popular as marijuana) abusers in Indonesia. If a marijuana addict suddenly wants to stop using marijuana, it can cause symptoms of “sakauâ€. To anticipate the symptoms of “sakauâ€, rehabilitation treatment can be taken, so that marijuana addicts can get comprehensive treatment. Determining the appropriate type of rehabilitation, can make it useful. Then knowing the last time abusers had consumption the marijuana, be expected to provide supporting information to determine the appropriate rehabilitation program for marijuana addicts. One technique in data mining that can be used to solve this problem is classification techniques. In this study using Radial Radial Basis Function Neural Network (RBFNN) with K-Means as the classification method. The steps taken included data normalization, K-Means to found the value of centers and spread for Gaussian activation function, training and testing RBFNN. This study using 627 marijuana abuser data which was published on the UCI Machine Learning in 2016. The results of the research showed the optimal parameters involves 7 hidden neurons and 100 as the maximum limit of K-Means iterations. By using these parameters, the classification result achieved accuracy of 35,908%.]]>
Copyrights © 2019
