TY  - THES
M1  - other
ID  - repoupi144381
Y1  - 2025/08/25/
PB  - Universitas Pendidikan Indonesia
KW  - Generalized Linear Model
KW  -  Machine Learning
KW  -  Neural Network
KW  -  Prediksi Produksi Perikanan
KW  -  Support Vector Machine

Fisheries Production Prediction
KW  -  Generalized Linear Model
KW  -  Machine Learning
KW  -  Neural Network
KW  -  Support Vector Machine
A1  - Ahmad Muzakki, -
AV  - restricted
N2  - Produksi perikanan tangkap yang akurat penting untuk mendukung pengelolaan sumber daya perikanan berkelanjutan di Pelabuhan Perikanan Nusantara (PPN) Palabuhanratu. Selama ini prediksi produksi dilakukan manual sehingga kurang efisien dan akurat. Penelitian ini menggunakan pendekatan machine learning untuk meningkatkan ketepatan estimasi, dengan data sekunder produksi perikanan tangkap periode 2021?2024 yang mencakup komoditas pelagis besar, pelagis kecil, demersal, serta data pendukung seperti jenis alat tangkap, ukuran kapal, dan jumlah trip. Tiga algoritma dibandingkan, yaitu Generalized Linear Model (GLM), Support Vector Machine (SVM), dan Neural Network (NN), melalui tahapan pembersihan dan transformasi data, pembagian data latih dan uji (80:20), pemodelan, serta evaluasi menggunakan Root Mean Squared Error (RMSE) dan confusion matrix. Hasil menunjukkan NN unggul dengan RMSE terendah (39.432) dan rata-rata accuracy, precision, recall tertinggi (96,21%), diikuti GLM (92,33%; RMSE 39.569) dan SVM (74,39%; RMSE 42.735). Berdasarkan hasil analisis, Neural Network (NN) terbukti unggul dan konsisten dalam memprediksi produksi perikanan tangkap di PPN Palabuhanratu, sehingga direkomendasikan sebagai model paling optimal pada penelitian ini. 


Accurate capture fisheries production is crucial to support sustainable fisheries resource management at the Palabuhanratu Archipelago Fisheries Port (PPN). Currently, production predictions are performed manually, making them less efficient and accurate. This study uses a machine learning approach to improve estimation accuracy, using secondary data on capture fisheries production for the 2021?2024 period, covering large pelagic, small pelagic, and demersal commodities, as well as supporting data such as fishing gear type, vessel size, and number of trips. Three algorithms were compared: the Generalized Linear Model (GLM), Support Vector Machine (SVM), and Neural Network (NN), through data cleaning and transformation, training and test data sharing (80:20), modeling, and evaluation using the Root Mean Squared Error (RMSE) and confusion matrix. The results showed that NN was superior with the lowest RMSE (39,432) and the highest average accuracy, precision, and recall (96.21%), followed by GLM (92.33%; RMSE 39,569) and SVM (74.39%; RMSE 42,735). Based on the analysis results, Neural Network (NN) was proven to be superior and consistent in predicting capture fisheries production in PPN Palabuhanratu, so it is recommended as the most optimal model in this study.
N1  - https://scholar.google.com/citations?hl=en&user=bH_TFzYAAAAJ&scilu=&scisig=AKwmTV4AAAAAaP4dRxInmGs5gCMSBsMZfp323rw&gmla=AKzYXQ3HOIdb4HBtotU0PzvLcPDazIYF5KJ7dXm1bXVLWoMPy8s-olS89vSetRarAdP37QAmBd2leex4w4R3KoQe2Sy6nPC2_gcPzmA&sciund=67022954651099208

ID SINTA Dosen Pembimbing :
AYANG ARMELITA ROSALIA: 6721849
WILLDAN APRIZAL ARIFIN: 6745746
TI  - PERBANDINGAN PERFORMA ALGORITMA MACHINE LEARNING UNTUK PREDIKSI PRODUKSI PERIKANAN TANGKAP DI PPN PALABUHANRATU
UR  - https://repository.upi.edu/
ER  -