Penyelesaian Permasalahan Penjadwalan Terintegrasi Di Rumah Sakit Menggunakan Algoritma Adaptive Threshold Iterated Local Search

Rifa'i, Dzaky Purnomo (2026) Penyelesaian Permasalahan Penjadwalan Terintegrasi Di Rumah Sakit Menggunakan Algoritma Adaptive Threshold Iterated Local Search. Other thesis, Institut Teknologi Sepuluh Nopember.

Warning
There is a more recent version of this item available.
[thumbnail of 5026221085-Undergraduate_Thesis.pdf] Text
5026221085-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only
Download (6MB)

Abstract

Rumah sakit menghadapi tantangan kompleks dalam mengelola penjadwalan sumber daya, khususnya yang melibatkan perencanaan operasi (surgical case planning), penjadwalan penerimaan pasien (patient admission scheduling), dan penugasan perawat (nurse-to-room assignment) secara terintegrasi. Permasalahan ini bersifat NP-Hard dengan ketergantungan yang tinggi antar variabel, di mana pendekatan konvensional yang memisahkan sub-masalah sering kali tidak efektif. Penelitian ini mengusulkan penerapan algoritma Adaptive Threshold Iterated Local Search (AT-ILS) untuk menyelesaikan permasalahan tersebut. Pendekatan dilakukan dalam dua fase: fase pertama menggunakan Progressive Acceptance Iterated Local Search (PA-ILS) untuk mencapai solusi feasible dengan memenuhi seluruh hard constraints, dan fase kedua menggunakan AT-ILS untuk meminimalkan pelanggaran soft constraints. Penelitian ini menggunakan dataset dari Integrated Healthcare Timetabling Competition (IHTC) 2024. Hasil pengujian menunjukkan bahwa algoritma yang dikembangkan mampu menghasilkan solusi feasible (0 pelanggaran hard constraint) pada seluruh 30 instance dataset uji. Dalam analisis kinerja, AT-ILS terbukti lebih unggul dibandingkan algoritma Hill Climbing dengan peningkatan kualitas solusi hingga 83,2%, serta kompetitif dibandingkan algoritma berbasis Quantum Learning pada instance skala kecil hingga menengah. Dalam validasi eksternal terhadap hasil kompetisi IHTC 2024, algoritma ini menempati peringkat ke-27 secara keseluruhan dengan keunggulan nol solusi unfeasible, membuktikan reliabilitasnya dibandingkan tim-tim lain yang gagal menghasilkan solusi layak. Selain itu, mekanisme ambang batas adaptif (adaptive threshold) terbukti efektif menyeimbangkan eksplorasi dan eksploitasi tanpa memerlukan pengaturan parameter manual yang rumit. Penelitian menyimpulkan bahwa AT-ILS merupakan metode yang robust untuk menyelesaikan masalah penjadwalan rumah sakit yang kompleks.
=====================================================================================================================================
Hospitals face complex challenges in managing resource scheduling, particularly involving integrated surgical case planning, patient admission scheduling, and nurse-to-room assignment. This problem is NP-Hard with high interdependencies between variables, where conventional approaches that separate sub-problems are often ineffective. This study proposes the application of the Adaptive Threshold Iterated Local Search (AT-ILS) algorithm to solve these problems. The approach is conducted in two phases: the first phase uses Progressive Acceptance Iterated Local Search (PA-ILS) to achieve feasible solutions by satisfying all hard constraints, and the second phase uses AT-ILS to minimize soft constraint violations. This research utilizes the dataset from the Integrated Healthcare Timetabling Competition (IHTC) 2024. Test results show that the developed algorithm is capable of producing feasible solutions (0 hard constraint violations) across all 30 test dataset instances. In performance analysis, AT-ILS proved superior to the Hill Climbing algorithm with solution quality improvements of up to 83.2%, and is competitive compared to Quantum Learning-based algorithms on small to medium-scale instances. In external validation against the IHTC 2024 competition results, this algorithm ranked 27th overall with the distinct advantage of zero unfeasible solutions, proving its reliability compared to other teams that failed to produce valid solutions. Furthermore, the adaptive threshold mechanism proved effective in balancing exploration and exploitation without requiring complex manual parameter tuning. The study concludes that AT-ILS is a robust method for solving complex hospital scheduling problems.

Item Type: Thesis (Other)
Uncontrolled Keywords: Adaptive Threshold Iterated Local Search, Penjadwalan Rumah Sakit, Optimasi Kombinatorial, Timetabling, Adaptive Threshold Iterated Local Search, Hospital Scheduling, Combinatorial Optimization, Timetabling
Subjects: T Technology > T Technology (General) > T57.84 Heuristic algorithms.
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User: Dzaky Purnomo Rifa`i
Date Deposited: 31 Jan 2026 04:32
Last Modified: 31 Jan 2026 04:32
URI: http://repository.its.ac.id/id/eprint/131349

Available Versions of this Item

Actions (login required)

View Item View Item