Implementasi Sistem Pengenalan Candi Kecil Di Yogyakarta Menggunakan Machine Learning Berbasis Cloud.

Abstract

In accordance with the national plan according to Presidential Regulation Number 58 of 2014 concerning the Spatial Plan for the Borobudur area, Borobudur is a national mainstay tourism area. As a result, data on visitors to Borobudur temple has increased greatly, the data for the last year of visitors to Borobudur temple every day can reach 55,000 people per day, especially during the national holiday season and school holidays, this has an impact on the condition of the temples and stupas which are increasingly exposed to friction between visitors. thus giving rise to new regulations regarding the daily visitor limit of only 1,250 per day. With the limitation of visitors to Borobudur temple, it is possible for tourists to change their tourist destinations to small temples around Yogyakarta which are less exposed by tourists, such as Sambisari temple, Gebang temple or Ijo temple and so on. one way to get to know more about the temples around Yogyakarta is to create a temple recognition system with the help of the implementation of cloud-based machine learning from nyckel.com, namely a platform as service for machine learning. from the test results on a dataset of 50 images with a distribution of 80 to 20. resulting in a confidence value of an average of 95% this number can prove that the temple recognition model with cloud-based machine learning can be used for temple recognition properly.

Keywords: Recognation System, Machine Learning, Cloud