An Application of Herbal Cough Medicine Selection Using Electre Method

Abstract

Nowdays, the development of traditional health services uses increasingly rapid concoctions. The use of herbal medicine as an alternative medicine by the Indonesian is not a new thing. Indonesia is a country that has a variety of plants that can be used as herbal medicines, including the use of herbal medicines for coughing. This study aims to design an application for selecting herbal cough medicines to make it easier for people to choose and recognize medicinal plants suitable for coughing. The algorithm used is Elimination and Choice Expressing Reality (ELECTRE) with the outranking concept using pairwise comparisons of alternatives based on each appropriate criterion. The research design used in building this system is structured UML and Software used by PHP and MySql for database processing. The results of this study are an application that is able to recommend herbs that are appropriate for each web-based cough category where electre will determine the ranking of the best alternatives of several types of herbal medicines according to the category or type of cough.

Klasifikasi Cacat Biji Kopi Menggunakan Metode Transfer Learning dengan Hyperparameter Tuning Gridsearch

“…Data biji kopi diproses dengan teknik computer vision melalui proses ekstraksi fitur untuk mendapatkan ciri-ciri biji kopi yang mengandung cacat dan yang tidak, kemudian algoritma belajar dari fitur tersebut untuk dilakukan proses pembelajaran selanjutnya. Algoritma machine learning melakukan pengelompokan terhadap data fitur yang telah dipelajari sebelumnya berdasarkan kategori biji kopi [6]. Pemilihan fitur merupakan tugas yang sangat penting dalam machine learning karena akan mempengaruhi kinerja dari model klasifikasi, akan tetapi pemilihan fitur juga dapat menjadi tugas yang rumit dan memerlukan pemikiran yang matang.…”