Abstract:Abstract: PT. Padasa Enam Utama, a palm oil plantation company, currently assesses the quality of Fresh Fruit Bunches (FFB) based solely on physical aspects. They use Microsoft Excel without a specialized application system,…
tem, which can lead to subjective assessments and the risk of fraud. This research aims to develop a predictive system for the quality of Fresh Fruit Bunches. Data collection was conducted using quantitative methods through direct observation and interviews with relevant parties. The study shows that the use of the K-Nearest Neighbor method provides the best accuracy with a more efficient calculation process. With this system, the company’s performance in making decisions regarding FFB quality is expected to improve. The system helps reduce human errors in quality assessments and offers visualizations that make it easier for users to understand the classification of production quality. Although the results are reliable, there is still room for further development, such as improving accuracy through more advanced data preprocessing techniques or using more complex machine learning models.
Keywords: Data Mining; K-Nearest Neighbor algorithm; Fresh Fruit Bunches (FFB); Python; Streamlit.
Abstrak: PT. Padasa Enam Utama, sebuah perusahaan perkebunan kelapa sawit, saat ini menilai kualitas produksi Tandan Buah Segar (TBS) berdasarkan aspek fisik saja. Mereka menggunakan Microsoft Excel tanpa sistem aplikasi khusus, yang dapat menyebabkan penilaian tidak objektif dan risiko kecurangan. Penelitian ini bertujuan untuk mengembangkan sistem prediksi mutu kualitas Tandan Buah Segar. Dalam pengumpulan data, digunakan metode kuantitatif melalui observasi langsung dan wawancara dengan pihak terkait. Penelitian menunjukkan bahwa penggunaan metode K-Nearest Neighbor memberikan akurasi terbaik dengan proses perhitungan yang lebih efisien. Dengan adanya sistem ini, diharapkan kinerja perusahaan dalam membuat keputusan terkait mutu produksi TBS dapat meningkat. Sistem ini membantu mengurangi kesalahan manusia dalam penilaian mutu TBS dan memberikan visualisasi yang memudahkan pengguna memahami klasifikasi mutu produksi. Meskipun telah memberikan hasil yang dapat diandalkan, masih ada ruang untuk pengembangan lebih lanjut, seperti peningkatan akurasi melalui teknik preprocessing data yang lebih canggih atau penggunaan model-machine learning yang lebih kompleks.
Kata kunci: Data Mining; Algoritma K-Nearest Neighbor; Tandan Buah Segar (TBS); Python; Streamlit
Abstract:Abstract: PT. Perkebunan Nusantara IV (Persero) Tinjowan is a state-owned enterprise unit operating in the palm oil plantation industry. In an effort to increase production, the plantation company aims to enhance production…
ion efficiency so that palm oil product prices become more competitive by determining harvest yields. The company's harvest yield estimation still uses a traditional approach based on the average bunch weight (BJR), considering the year of planting and the planting area. However, this technique is less effective and efficient. If the company's estimates are incorrect, it can potentially lead to losses or increased production budgets. The purpose of this research is to obtain a comparison of forecast results for the next month with the best alpha accuracy measure using the Single Exponential Smoothing method. The research method used is based on qualitative and quantitative data from interviews and observations of data from December 2022 to November 2023. The forecasting results with an alpha accuracy value of 0.9 using the Single Exponential Smoothing method show a forecast for December 2023 of 13,835.57844 with a percentage error rate (MAPE) of 9.28%, MAD of 1,021,423.6, and MSE of 174,130,366.
Keywords: forecasting; palm oil; single exponential smoothing
Abstrak: PT. Perkebunan Nusantara IV (Persero) Tinjowan sebagai unit usaha BUMN yang beroperasi di industri perkebunan kelapa sawit. Dalam upaya meningkatkan produksi, perusahaan perkebunan, peningkatan efisiensi produksi agar harga produk sawit lebih kompetitif dengan menentukan hasil panen. Pada perkiraan hasil panen perusahaan masih menerapkan pendekatan tradisional berdasarkan berat janjangan rata-rata (BJR) dengan mempertimbangkan tahun tanam dan luas area tanaman. Namun, tehnik ini kurang efektif dan efisien, Apabila perkiraan yang dibuat perusahaan salah berpotensi menyebabkan kerugian atau peningkatan anggaran produksi. Tujuan dari penelitian ini untuk mendapatkan perbandingan hasil peramalan pada bulan berikutnya dengan ukuran akurasi alpha terbaik menggunakan metode Single Exponential Smoothing. Metode yang digunakan dalam penelitian ini adalah berdasarkan data kualitatif dan kuantitatif dari hasil wawancara dan observasi data bulan desember 2022 - november 2023. Hasil peramalan dari ukuran akurasi nilai alpha 0.9 menggunakan metode Single Exponential Smoothing diperoleh peramalan untuk periode desember 2023 sebesar 13.835578,44 dengan tingkat persentase error MAPE sebesar 9,28%, MAD 1021423.6 dan MSE 174130366.
Kata Kunci: kelapa sawit; peramalan; single exponential smoothing
Abstract:This study expects to determine the impact of cow compost and urea manure to increase yields and ideal plant development. This exploration was completed at the Medan City Food Security Administration, Jl. Delightful Kramat/Selambo…
at/Selambo Ujung. From May to August 2023. This study utilized a factorial randomized bunch plan (RBD) which comprised of 2 treatment factors with the image S which comprised of 3 levels, in particular; ( S1) 1 kg/plot (S2) 2 kg/plot and (S3) 3 kg/plot. The subsequent factorial is the U image which consists of 3 levels to be specific; (U1) 30 g/plot (U2) 40 g/plot and (P3) 50 g/plot. With the goal that 9 medicines were obtained. Every treatment was rehashed multiple times. The perception information was then examined with the F trial of change at the 5% level, in the event that the thing that matters was huge, the test was gone on with the Duncan test. From the aftereffects of the examination and conversation completed, the accompanying ends were obtained: 1). The impact of cow fertilizer application affected plant level, number of leaves, number of blossoms, number of cases and unit weight of bean plants (Phaseolus vulgaris L.) 2). The impact of applying urea manure meaningfully affected plant level, significantly affected the quantity of leaves, fundamentally affected the quantity of blossoms and affected the quantity of cases and unit weight of bean plants (Phaseolus vulgaris L.) 3). The communication between the treatment of cow excrement and urea manure essentially affected plant level, number of blossoms, number of cases and unit weight, but not altogether unique on the quantity of leaves of bean plants (Phaseolus vulgaris L.)
Abstract:This study aims to analyze the Fresh Fruit Bunch Boiling Process at the Sterlizer Station in Palm Oil Processing at PT. Sumber Bumi Sawit Jadi Jaya. In writing this report, data collection is very important. The data collection…
lection methods used by the author are as follows: Documents or References Data collection through written and electronic documents from PT. Sumber Bumi Sawit Jadi Jaya. Field Observation. The method is carried out by direct observation in the company environment. Literature Review. In addition to field observations, data is also taken from several literatures related to this report. The location of the field work practice was held at PT Sumber Bumi Sawit Jadi Jaya which is located on Jalan Besar Mandoge-Kisaran. And the field work practice lasted for 1 month. Starting from September 1st – October 1st, 2023. The conclusions obtained from the results of the practical work are as follows: PT. Sumber Bumi Sawit Jadi Jaya is a factory that processes fresh oil palm fruit bunches into CPO and palm kernel. The processing capacity at PT. Sumber Bumi Sawit Jadi Jaya is 30 Tons of FFB / Hour The organizational structure at PT. Sumber Bumi Sawit Jadi Jaya is a functional organizational structure. The results of processing the percentage of cardiovascular load obtained that all boiler workers needed repairs, and the workload category of the four workers was influenced by age factors