Pengembangan Sistem Pendukung Keputusan Pemilihan Menu Makanan Berbasis Website Dengan Greedy Algorithm Bagi Penderita Penyakit Kronis Dan Tidak Menular

Primadhani, Dicky Febri (2026) Pengembangan Sistem Pendukung Keputusan Pemilihan Menu Makanan Berbasis Website Dengan Greedy Algorithm Bagi Penderita Penyakit Kronis Dan Tidak Menular. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026221036-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Penyakit kronis dan tidak menular seperti diabetes melitus tipe 2, hipertensi, penyakit kardiovaskular, hiperkolesterolemia, dan penyakit ginjal kronis memerlukan pengelolaan pola makan yang ketat. Pemilihan menu harian yang memenuhi batasan nutrisi klinis secara simultan, terutama pada kasus komplikasi multi-penyakit, merupakan permasalahan optimasi yang belum banyak diselesaikan secara komputasional. Penelitian ini mengembangkan Sistem Pendukung Keputusan berbasis web bernama ChronoBite yang membantu pengguna menentukan menu makanan harian sesuai kebutuhan nutrisi dan kondisi kesehatannya. Sistem memanfaatkan Greedy Algorithm sebagai mekanisme pemilihan menu secara bertahap pada setiap waktu makan, dengan memastikan setiap rekomendasi berada dalam batasan nutrisi klinis yang ditetapkan. Data nutrisi bersumber dari USDA FoodData Central dan aturan batasan nutrisi disusun berdasarkan pedoman diet klinis resmi per penyakit. Pada kasus multi-penyakit, seluruh batasan nutrisi yang relevan digabungkan menggunakan pendekatan intersection constraint. Evaluasi terhadap 26 skenario profil penyakit menunjukkan rata-rata Constraint Satisfaction Rate (CSR) sebesar 48,54%, dengan kinerja terbaik pada profil Normal (71,43%) dan tersulit pada profil DM2 + Hiperkolesterolemia + CKD (33,33%). Rata-rata deviasi nutrisi berada di 48,32%, di mana makronutrien utama berhasil dikendalikan dengan baik sementara beberapa mikronutrien menunjukkan deviasi tinggi akibat keterbatasan dataset. Validasi Ahli Gizi terhadap 13 skenario menghasilkan skor rata-rata 3,26 dari 5 (Cukup Sesuai), dengan catatan perbaikan terkait keberagaman komponen sajian. Seluruh test case pengujian Black Box berhasil lulus. Dari sisi efisiensi, sistem menghasilkan rekomendasi dalam rata-rata 1,33 detik per eksekusi, menjadikannya sesuai untuk aplikasi web yang menuntut respons real-time.
================================================================================================================================
Chronic non-communicable diseases such as type 2 diabetes mellitus, hypertension, cardiovascular disease, hypercholesterolemia, and chronic kidney disease require strict dietary management. Selecting daily menus that simultaneously satisfy clinical nutritional constraints, particularly in multi-disease complication cases, represents a computational optimization problem that remains largely unsolved. This study develops a web-based Decision Support System named ChronoBite that assists users in determining daily meal plans aligned with their nutritional needs and health conditions. The system employs a Greedy Algorithm as a stepwise menu selection mechanism for each meal time, ensuring every recommendation remains within predefined clinical nutritional boundaries. Nutritional data is sourced from the USDA FoodData Central, and nutritional constraint rules are derived from official clinical dietary guidelines for each disease. For users with multiple health conditions, all relevant nutritional constraints are combined using an intersection constraint approach. Evaluation across 26 disease profile scenarios demonstrated an average Constraint Satisfaction Rate (CSR) of 48.54%, with the best performance on the Normal profile (71.43%) and the most challenging on the DM2 + Hypercholesterolemia + CKD profile (33.33%). Average nutrient deviation stood at 48.32%, where primary macronutrients were well-controlled while several micronutrients exhibited high deviation due to dataset limitations. Nutritionist validation across 13 case scenarios yielded an average score of 3.26 out of 5 (Sufficiently Appropriate), with noted improvements regarding meal component diversity. All Black Box test cases passed successfully. In terms of computational efficiency, the system generates recommendations in an average of 1.33 seconds per-execution, making it well-suited for web applications demanding real-time response.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Sistem Pendukung Keputusan, Rekomendasi Menu Makanan, Diet Klinis, Penyakit Tidak Menular, Decision Support System, Greedy Algorithm, Dietary Menu Recommendation, Clinical Diet, Non-Communicable Disease
Subjects:	T Technology > T Technology (General) > T58.62 Decision support systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Dicky Febri Primadhani
Date Deposited:	28 Jul 2026 02:21
Last Modified:	28 Jul 2026 02:21
URI:	http://repository.its.ac.id/id/eprint/138255

Actions (login required)

View Item