Implementasi Metode K-nearest Neighbor Dalam Menentukan Kualitas Massa Batuan

Rahmadanti, Siti (2020) Implementasi Metode K-nearest Neighbor Dalam Menentukan Kualitas Massa Batuan. Other thesis, Universitas Islam Riau.

Preview

Text 153510492.pdf - Submitted Version Download (1MB) | Preview

Abstract

During this time the state of the structure of rock mass in nature which tends to be different is controlled by the appearance of the geological structure, the field of discontinuity, and the field of bedding or burly. The consequence that is often faced in open mines is deformation behavior, and poor condition of rock mass structure. Therefore, the evaluation of slope stability is an important part of preventing disruption to production smoothness and fatal disasters. Meanwhile, in relation to rock engineering, Rock mass classification means collecting data and classifying rock outcrops based on parameters that are believed to reflect the rock mass behavior. The main use of the rock mass classification system is to assess various technical properties of or associated with rock mass with a predetermined weight value. In this study a system was developed using the k-Nearest Neighbor method with 9 parameters of rock mass classification (Rock Mass Classification) with the aim of determining the level of accuracy of rock mass quality and making it easier to calculate the value of the parameters so as to produce accurate results. KNN is one of the best and most widely used classification data mining algorithms. K in KNN is a variable of the number of nearest neighbors that will taken for the classification process. This study produced the best K in the experiment K = 14 with an accuracy of 75%. K = 14 is the most optimal k value among KNN classification experiments using values K = 2 to K = 18. This system is implemented with a web programming language with PHP (Hypertext Preprocessor) and uses a MySQL database.

Actions (login required)

View Item