RANCANG BANGUN SISTEM PREDIKSI TINGKAT KERENTANAN GAGAL PANEN PADA PETANI PADI DI KECAMATAN BUAY MADANG TIMUR MENGGUNAKAN ALGORITMA K-NEAREST NEIGHBOR (KNN)

Silvia, Amanda (2026) RANCANG BANGUN SISTEM PREDIKSI TINGKAT KERENTANAN GAGAL PANEN PADA PETANI PADI DI KECAMATAN BUAY MADANG TIMUR MENGGUNAKAN ALGORITMA K-NEAREST NEIGHBOR (KNN). Diploma thesis, UIN RADEN INTAN LAMPUNG.

	PDF Download (5MB)
	PDF Restricted to Repository staff only Download (5MB)

Abstract

Kecamatan Buay Madang Timur merupakan salah satu wilayah penghasil padi di Kabupaten OKU Timur yang mengalami fluktuasi produksi akibat faktor iklim, serangan hama, dan kondisi lingkungan. Selain itu, belum tersedia sistem berbasis data untuk mengidentifikasi tingkat kerentanan gagal panen sehingga penilaian kondisi lahan dan tanaman masih dilakukan berdasarkan pengalaman dan pengamatan subjektif. Penelitian ini bertujuan merancang dan membangun Sistem berbasis web untuk memprediksi tingkat kerentanan gagal panen menggunakan algoritma K-Nearest Neighbor (KNN). Penelitian menggunakan metode Research and Development (R&D) dengan model pengembangan Waterfall. Sistem dikembangkan menggunakan PHP Native, dan MySQL, sedangkan algoritma KNN digunakan untuk mengklasifikasikan tingkat kerentanan gagal panen berdasarkan variabel luas lahan, varietas padi, sistem irigasi, penggunaan pupuk, penggunaan pestisida, intensitas serangan hama, frekuensi tanam per tahun, dan riwayat gagal panen. Pengujian dilakukan menggunakan Black Box Testing dan System Usability Scale (SUS). Hasil penelitian menunjukkan bahwa sistem berhasil dikembangkan dengan fitur pengelolaan data historis, prediksi, pengujian akurasi, dan penyajian hasil prediksi. Seluruh fungsi sistem berjalan dengan baik dan hasil pengujian usability menunjukkan sistem layak digunakan dalam memprediksi tingkat kerentanan gagal panen. Kata Kunci: Sistem, Prediksi Gagal Panen, K-Nearest Neighbor (KNN), Waterfall, Research and Development (R&D). Buay Madang Timur District is one of the major rice-producing areas in OKU Timur Regency that experiences fluctuations in rice production due to climatic conditions, pest infestations, and environmental factors. In addition, there is no data-driven system available to identify the level of crop failure vulnerability, causing the assessment of land and crop conditions to rely primarily on farmers' experience and subjective observations. This study aims to design and develop a web-based system for predicting the level of crop failure vulnerability using the K-Nearest Neighbor (KNN) algorithm. The research employed the Research and Development (R&D) method with the Waterfall development model. The system was developed using Native PHP and MySQL, while the KNN algorithm was applied to classify the level of crop failure vulnerability based on land area, rice variety, irrigation system, fertilizer usage, pesticide usage, pest infestation intensity, planting frequency per year, and crop failure history. System evaluation was conducted using Black Box Testing and the System Usability Scale (SUS). The results indicate that the system was successfully developed with features for managing historical data, performing vulnerability predictions, evaluating prediction accuracy, and presenting prediction results. All system functions operated properly, and the usability testing results demonstrated that the system is suitable for predicting the level of crop failure vulnerability. Keywords: System, Crop Failure Prediction, K-Nearest Neighbor (KNN), Waterfall, Research and Development (R&D).

Actions (login required)

View Item