Putri, Sephia Dwi Arma (2023) Implementasi Algoritma Multiple Linear Regression Dan Classification Naïve Bayes Dalam Analisis Prediktif Perubahan Nilai Profit Berdasarkan Pengiriman Outgoing Pada J&T Cargo Kalipare. Undergraduate thesis, Fakultas Teknologi Informasi Universitas Merdeka Malang.

	Text COVER.pdf Download (3MB)
	Text BAB I.pdf Download (1MB)
	Text BAB II.pdf Restricted to Repository staff only Download (1MB)
	Text BAB III.pdf Restricted to Repository staff only Download (550kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (1MB)
	Text BAB V.pdf Restricted to Repository staff only Download (185kB)
	Text DAFTAR PUSTAKA.pdf Download (329kB)
	Text HASIL CEK PLAGIASI.pdf Download (452kB)

Abstract

The 78% increase in e-commerce users is positively correlated to the increase in logistics service users (expeditions). Not only shipping goods from e- commerce sales, but expeditions have finally expanded to serve the delivery of goods not only through purchase transactions but the distribution of goods. J&T Cargo Kalipare serves the delivery of goods with large capacity and volume at an affordable cost. J&T Cargo Kalipare has a fairly complex problem, where based on the SOP, each outlet is required to meet the shipping target (tonnage) of 1000 Kg every month, with a minimum achievement of 20%. When the target is not achieved, a fine of Rp. 200,000 and an increase in the tonnage target will be imposed. This makes the operational burden increase and profit instability. Therefore, the author conducts research on how to anticipate the non achievement of the target so that the outlet is not subject to fines and profit becomes stable. The use of these two algorithms has their respective functions. The Naïve Bayesalgorithm is used to classify customers into Industrial or Individual classes, with the use of this algorithm considered suitable for the data shown with a Precision of 0.756 and Recall of 0.759. The result is that 69.33% of industrial categories contribute a lot to delivery and the remaining 30.67% are individual categories. The second algorithm is Multiple Linear Regression which functions to predict profits according to category class. Based on the calculation of significance F against alpha shows 1.3855e-77 so the algorithm is considered suitable for the data, by producing predictions from October to December experiencing an increase in profits in the individual category and a decrease in the industrial category. So that there are recommendations that can be made, including royalty rewards for loyal customers, brand awareness to attract new customers, and business cooperation with UMKMs and other business people around

Actions (login required)

View Item