Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Jurnal Informatika Upgris

PENERAPAN METODE NAÏVE BAYES DALAM PEMERIKSAAN KESEHATAN Nurtriana Hidayati; Victor Gayuh Utomo; nur wakhidah
Jurnal Informatika Upgris Vol 5, No 2: Desember (2019)
Publisher : Universitas PGRI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26877/jiu.v5i2.4313

Abstract

Maintaining the safety and health of its workers by conducting daily health checks. The medical examination is carried out by a doctor or nurse at work. Workers who were declared unhealthy were not allowed to work on that day, in several cities, so that inspection standards were maintained, a web-based system was implemented that implemented a rule-based expert system. Naive Bayes algorithm which is included as an algorithm for classification based on data mining and machine learning activities. The initial process of Naive Bayes begins by studying existing data (complete with labels) through the training process. The results of this training are actually the probability of an outcome based on the variables used. Naive Bayes was chosen because it can study the results of past examinations and use them to make the next decision. This is felt to be a solution to the problems of previous system deficiencies that require the experience and expertise of health workers who are not necessarily equal.
PREDIKSI HARGA SAHAM DENGAN SVM (SUPPORT VECTOR MACHINE) DAN PEMILIHAN FITUR F-SCORE Victor Gayuh Utomo; Nur Wakhidah; Astrid Novita Putri
Jurnal Informatika Upgris Vol 6, No 1: JUNI (2020)
Publisher : Universitas PGRI Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26877/jiu.v6i1.5306

Abstract

Dalam pasar modal terdapat dua buah metode yang digunakan investor untuk membuat prediksi harga saham yaitu analisa fundamental dan analisa teknikal. Penelitian ini melakukan prediksi harga saham berdasarkan analisa teknikal dengan menggunakan indikator teknikal sebagai fiturnya. Penelitian menggunakan metode Support Vector Machine (SVM) untuk prediksi dan melakukan pemilihan fitur (feature selection) dengan metode F-Score.Penelitian telah menyelesaikan pengembangan prototype yang dibutuhkan untuk melakukan prediksi saham. Pengumpulan data saham dari Bursa Efek Indonesia telah dilaksanakan dan proses prediksi yang dibutuhkan juga telah dilaksanakan.Metode F-Score membutuhkan proses yang jauh lebih sedikit daripada proses maksimal yang mungkin dilakukan untuk mencari fitur terbaik terutama setelah jumlah fitur melebihi 5. Metode F-Score juga memberikan hasil 70% F-Score dan 71% akurasi. Hasil ini hanya 3% lebih buruk daripada pilihan fitur terbaik yang mungkin dicapai