Hasil Pencarian  ::  Simpan CSV :: Kembali

"A ship usually performs maneuvers under the influence of external forces and moments, such as wind, waves, and current. Therefore, it is important to understand the maneuvering behavior of ships under the action of external forces. This paper discusses the turning maneuvers of an Indonesian roro ferry under the combined influence of constant wind and regular waves using the mathematical modelling group (MMG). The ship’s position relative to the wave trough is added to the original MMG model to estimate the exciting forces and moment induced by the waves. The results of a numerical simulation show that the effect of wave height on turning ability is more significant for a small wavelength; this effect decreases as the wavelength increases. The effect of wavelength on the sway force and yaw moment is more significant compared with its effect on the surge force. The ship’s initial position relative to the wave trough does not have a significant effect on the turning characteristic and it can be neglected for the present study’s subject ship. Overall, the results of the present work compare well with published data."

"Produk susu bubuk balita yang beragam membuat konsumen memiliki banyak pilihan sehingga penting bagi produsen menjaga loyalitas pelanggan yang telah ada dengan memahami perilaku churn pelanggan. Churn pelanggan didefinisikan sebagai kecenderungan pelanggan untuk berhenti melakukan bisnis dengan sebuah perusahaan. Penelitian ini berfokus memprediksi pola churn pelanggan sehingga perusahaan dapat menentukan strategi untuk mengurangi churn. Penelitian ini membahas mengenai prediksi churn pelanggan berdasarkan segmen produk susu bubuk balita menggunakan model Length, Recency, Frequency, Monetary (LRFM). Responden penelitian ini adalah pelanggan PT. XYZ yang pernah bertransaksi untuk produk susu bubuk balita kelas premium (susu A) dan segmen biasa (susu B) selama periode tahun 2021. Variabel pada penelitian ini meliputi variabel LRFM dan CLV yang dibentuk dengan pembobotan variabel LRFM. Pertama metode Fuzzy C-Means Clustering digunakan untuk melakukan pelabelan target pelanggan selanjutnya metode klasifikasi K-Nearest Neighbor (KNN) digunakan untuk memprediksi churn. Hasilnya terdapat tiga kelompok pelanggan untuk masing-masing susu A dan susu B. Pelabelan yang dihasilkan yaitu pelanggan churn dengan nilai CLV rendah, potential to churn dengan nilai CLV menengah, dan loyal dengan nilai CLV tinggi. Susu B menunjukkan jumlah pelanggan churn sebesar 43,4% lebih banyak dibandingkan susu A sebanyak 34%. Tahapan akhir penelitian ini adalah menganalisis kinerja metode KNN berdasarkan nilai akurasi, recall, dan f1-score terhadap kedua susu A dan susu B. Hasil dari tugas akhir ini menunjukkan bahwa kinerja metode KNN bergantung pada pemilihan jumlah tetangga terdekat dan proporsi pemisahan data.

---

The variety of powdered toddler milk products gives consumers many choices, so producers need to maintain the loyalty of existing customers by understanding customer churn behaviour. Customer churn is defined as the tendency of a customer to stop doing business with a company. This study focuses on predicting customer churn patterns so companies can determine strategies to reduce churn. This study discusses the prediction of customer churn based on the segment of toddler powdered milk products using the Length, Recency, Frequency, Monetary (LRFM) model. The respondent of this research are the customers of PT. XYZ who have transacted for premium segment powdered milk products for toddlers (milk A) and ordinary segment (milk B) during 2021. Variables in the data include LRFM and CLV variables which are formed by weighting the LRFM variable. At first, Fuzzy C-Means Clustering algorithm was applied for labelling target customer and then, K-Nearest Neighbor (KNN) Classifier as churn prediction was used. As a result, there are three groups of customers for each milk A and milk B. The resulting labels are the churn customer group with low CLV value, potential to churn group with medium CLV, and loyal customer group with high CLV value. Milk B shows the number of customers churn by 43,4% more than milk A as much as 34%. In the final stage of this research, the author analyze the performance of the KNN method based on the value of accuracy, recall, and f1-score for both milk A and milk B. The results of this final project show that the performance of the KNN method depends on the selection of the number of nearest neighbors and the proportion of data splitting used."