Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal RABIT: Jurnal Teknologi dan Sistem Informasi Univrab Jurnal Penelitian Pendidikan IPA (JPPIPA)
Heri Subangkit
Universitas Amikom Purwokerto
Agriculture, Biological Sciences & Forestry Biochemistry, Genetics & Molecular Biology Chemical Engineering, Chemistry & Bioengineering Chemistry Computer Science & IT Education Engineering Materials Science & Nanotechnology Physics

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

Found 2 Documents
Search

 1 
Lecturer Performance Prediction Based on Student Evaluation Data Using a Hybrid K-Means and Random Forest Model Heri Subangkit; Taqwa Hariguna; Dhanar Intan Surya Saputra
Jurnal Penelitian Pendidikan IPA Vol 12 No 1 (2026)
Publisher : Postgraduate, University of Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29303/jppipa.v12i1.14163

Abstract

Using a quantitative correlational design, this predictive research was based on secondary EDOM data. The first episode of the school year 2024/2025 served as the data collection period. The target population of this research are the lecturer subjected to students’ evaluations from Universitas Al-Irsyad Cilacap. After processing the data and cleaning and aggregating, a total of 594 records of the lecturer were analyzed with a census technique. K-Means was used to detect the presence of latent patterns of performance in the teaching, professional, personality and social dimensions of the lecturer. The Random Forest model was used to predict the performance category of the lecturer from both the baseline and hybrid models. The results of the study showed that the hybrid models were able to predict with a high measure of accuracy, and of the two, the hybrid model was the most robust when compared to the baseline model with a manual high-defined grouping of performance levels. The baseline model was able to completely and perfectly classify the group, the hybrid model with high performance was able to analyze the data in a general way, revealing a structure of performance that was hidden in the data. This means that, there is greater analytical value to the data. This analysis of EDOM data is of high analytical value. The developing of the hybrid model of lecturer performance analysis provides a positive contribution in data-driven quality assurance and decision-making to higher education. Objectives were met.
ANALISIS PROFILING KINERJA DOSEN BERBASIS ALGORITMA CLUSTERING: STUDI KASUS DATA EVALUASI MAHASISWA (EDOM) Heri Subangkit; Taqwa Hariguna; Dhanar Intan Surya Saputra
Rabit : Jurnal Teknologi dan Sistem Informasi Univrab Vol 11 No 1 (2026): Januari
Publisher : LPPM Universitas Abdurrab

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36341/rabit.v11i1.7455

Abstract

Lecturer quality is a key factor determining the success of a higher education institution. Although EDOM assessment is crucial for performance, large data processing often does not provide sufficient strategic information. This study utilizes a no-learning method to profile lecturer performance. K-Means and Fuzzy C-Means (FCM) are two clustering algorithms that are compared with four competency variables. These variables are pedagogical, professional, social, and personality. The results show that there are three ideal clusters (k = 3) which are categorized as "Very Good", "Good", and "Fair" performance groups, respectively. The K-Means algorithm produces a Silhouette Score of 0.507. However, FCM is more flexible in determining the number of data transition members required. The profiling results show that pedagogical competence is the variable with the lowest score in the "Fair" cluster. The findings of this study suggest that, to improve the quality of educational services, lecturers in the cluster need to undergo training related to teaching methods.
Co-Authors Dhanar Intan Surya Saputra Taqwa Hariguna