Pengembangan Aplikasi Prediksi Performa Akademik dan Sistem Pakar Mahasiswa Kedokteran Berbasis Web

Fano, Naufal Firjatulloh (2025) Pengembangan Aplikasi Prediksi Performa Akademik dan Sistem Pakar Mahasiswa Kedokteran Berbasis Web. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026211011-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Profesi Kedokteran merupakan suatu pekerjaan kedokteran yang dilaksanakan berdasarkan suatu keilmuan dan kompetensi yang diperoleh melalui pendidikan berjenjang. Sebagai upaya menjamin kualitas pendidikan profesi dokter di Indonesia, pemerintah menyelenggarakan ujian kompetensi akhir berdasarkan dengan Standar Kompetensi Dokter Indonesia (SKDI). Terdapat mahasiswa dari program studi Pendidikan Dokter Fakultas Kesehatan Universitas Muhammadiyah Surabaya (FKUM) yang tidak lulus pada percobaan pertama ujian kompetensi akhir, sehingga diharuskan untuk mengulang kembali ujian kompetensi akhir pada periode selanjutnya. Model prediksi dirancang menggunakan algoritma klasifikasi XGBoost karena memiliki keunggulan dalam hal performa klasifikasi set data tidak seimbang. Sementara itu, pembuatan sistem pakar dilakukan dengan knowledge base yang diperoleh dari dosen Medical Education Unit program Pendidikan Dokter Fakultas Kesehatan Universitas Muhammadiyah Surabaya (FKUM) melalui wawancara. Pengetahuan dari pakar direpresentasikan dalam bentuk Rule-Based System dengan total tujuh aturan. Inferensi sistem pakar menggunakan mekanisme forward chaining, dan rekomendasi sistem pakar akan diberikan berdasarkan hasil model prediksi. Model prediksi terbaik didapatkan pada model semester 4 dengan K=5 menghasilkan F1 score terhadap data validation sebesar 93.63% dan F1 score generalization sebesar 80.00%. Model prediksi dan sistem pakar diimplementasikan kedalam aplikasi berbasis web menggunakan Django web framework dan Tailwind CSS. Hasil evaluasi aplikasi menunjukkan skor SUS sebesar 75 untuk model prediksi dan 77.5 untuk sistem pakar. Penelitian ini diharapkan dapat membantu dosen FKUM untuk melakukan identifikasi dini ketidaklulusan mahasiswa dan memberikan rekomendasi perlakuan akademik dalam mempersiapkan ujian kompetensi akhir.
==========================================================================================================
The medical profession is a job requires scientific knowledge and competence, which are obtained through tiered education. To ensure the quality of medical professional education in Indonesia, the government holds a final competency exam based on Standar Kompetensi Dokter Indonesia (SKDI). Some students from medical education program at Fakultas Kedokteran Universitas Muhammadiyah Surabaya (FKUM) fail the final competency exam on their first attempt, and are required to retake it in the next period. Therefore, this study aims to develop a web-based academic performance prediction application and expert system. The prediction model was designed using the XGBoost classification algorithm because it has advantages in terms of classification performance of unbalanced data sets. Meanwhile, the creation of the expert system was carried out with a knowledge base obtained from lecturers of medical education program at Fakultas Kedokteran Universitas Muhammadiyah Surabaya (FKUM) through interviews. Knowledge from the expert is represented in the form of a Rule-Based System with a total of seven rules. Expert system inference uses a forward chaining mechanism, and expert system recommendations will be given based on the results of the prediction model. The best prediction model obtained by 4th semester model with K = 5 resulting in F1 score against validation data of 93.63% and F1 score generalization of 80.00%. The prediction model and expert system are implemented into a web-based application using Django web framework and Tailwind CSS. The application evaluation results showed a SUS score of 75 for the prediction model and 77.5 for the expert system. This research is expected to help FKUM lecturers to conduct early identification of student failure and provide academic treatment recommendations in preparing the final competency exam.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Ujian Kompetensi Akhir Kedokteran, Educational Data Mining, Model Prediksi, XGBoost, Sistem Pakar, Medical Final Competency Test, Educational Data Mining, Prediction Model, XGBoost, Expert System.
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.76.E95 Expert systems Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Naufal Firjatulloh Fano
Date Deposited:	31 Jan 2025 08:01
Last Modified:	31 Jan 2025 08:12
URI:	http://repository.its.ac.id/id/eprint/117301

Actions (login required)

View Item