Article Per Year (5 Year)
p-Index From 2020 - 2025
0.408
P-Index
This Author published in this journals
All Journal Prosiding Seminar Nasional Sisfotek (Sistem Informasi dan Teknologi Informasi)
Imam T. Umagapi
Unknown Affiliation
Computer Science & IT

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Prediksi Kelulusan Pegawai Pemerintah Dengan Perjanjian Kerja Guru Menggunakan Metode Naive Bayes Basirung Umaternate; Imam T. Umagapi; Yuyun, Yuyun; Hazriani, Hazriani
Prosiding SISFOTEK Vol 7 No 1 (2023): SISFOTEK VII 2023
Publisher : Ikatan Ahli Informatika Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Recruitment and selection of ASN PNS and PPPK candidates to date has quite a high number of enthusiasts, and the selection process using the Computer Assisted Test (CAT) is the main requirement for passing the PNS or PPPK selection. Therefore, the researcher made this research using pre-existing data or called training data. Researchers used the Naïve Bayes method with 9 mutually independent attributes to determine graduation. Researchers also used Microsoft Excel and Weka supporting applications to test the accuracy of the Naïve Bayes method. Tests were carried out with 212 datasets consisting of 170 training data/training data and 42 test data/testing. The results of the Accuracy, Recall, and Precesion tests determine the graduation of Government Employees with Work Agreements (PPPK) for Teachers in Morotai Island Regency, 100% Accuracy, 100% Precision, 100% Recall, and the Area Under ROC (AUC) value is 1, which means 100% below The curve shows that the performance of the Naïve Bayes algorithm in capitalizing the classification of graduation data for Government Employees with Employment Agreements (PPPK) for Teachers in Puau Morotai Regency is very good.
Uji Kinerja K-Means Clustering Menggunakan Davies-Bouldin Index Pada Pengelompokan Data Prestasi Siswa Imam T. Umagapi; Basirung Umaternate; Hazriani, Hazriani; Yuyun, Yuyun
Prosiding SISFOTEK Vol 7 No 1 (2023): SISFOTEK VII 2023
Publisher : Ikatan Ahli Informatika Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

This research investigates how the values of clustered datasets, both normalized and non-normalized, influence the computation of Euclidean distance in the K-means algorithm. Additionally, it examines the impact of varying cluster quantities, identified through the elbow method, on the evaluation of the Davies-Bouldin Index (DBI). A dataset comprising 174 records undergoes mining using the CRISP-DM (Cross-Industry Standard Process for Data Mining) approach. In the data preparation phase, the min-max algorithm is applied to ensure that attribute values within the dataset are not diminished relative to each other. Concerning the selection of an optimal K value, the elbow method is employed. In this investigation, two K values exhibit significant mean reduction: the fourth and third cluster quantities. The DBI results for 3 clusters show a smaller value of 0.9250 compared to the DBI result for 4 clusters, which is 1.1584. The fundamental principle of evaluating the Davies-Bouldin Index is that a smaller DBI value (approaching zero but not reaching the minimum) indicates a better cluster. These findings contribute to a better understanding of the evaluation techniques involving the elbow method and Davies-Bouldin Index in clustering analysis and offer insights into the relationship between determining cluster quantities and clustering performance.
Co-Authors Basirung Umaternate Hazriani, Hazriani Yuyun, Yuyun