Implementation of Automatic Clustering Algorithm and Fuzzy Time Series in Motorcycle Sales Forecasting

Abstract

Abstract. Accurate forecasting for the sale of a product depends on the forecasting method used. The purpose of this research is to build motorcycle sales forecasting application using Fuzzy Time Series method combined with interval determination using automatic clustering algorithm. Forecasting is done using the sales data of motorcycle sales in the last ten years. Then the error rate of forecasting is measured using Means Percentage Error (MPE) and Means Absolute Percentage Error (MAPE). The results of forecasting in the one-year period obtained in this study are included in good accuracy. IntroductionProfit is one of criterions used to measure the performance of a company [1]. The company's rate profitability is directly proportional to the level of sales [2], if the company profits as might be expected, then the company can expand its business. One of the fundamental problems that often occur in making sales plans difficult to realization is the mistake in making the right sales prediction [3]. If the sales prediction is too large (Over Production) then the cost of production will swell and all asset which is invested into less efficient. Conversly, if the sales prediction is made too small (Under Production) then the company will face out of stock (Stock Out), so the customer was forced to wait too long for the desired product [4]. Then it requires an activity to estimate the amount of sales of goods and services by the manufacturer and distributor in a certain time period i.e. sales forecasting. Right forecasting has several important criteria, including accuracy, cost and convenience. The selection of appropriate forecasting methods is a matter of concern for accurate forecasting [5].Sales forecasting will be influenced by previous sales data, then a suitable forecasting method used in sales forecasting is the time series method included the fuzzy time series method. This method is a fairly new method compared to other classical time series forecasting methods, such as Auto Regresive (AR), Moving Average (MA), ARMA, ARIMA, and so on. This method was used to forecast student enrollment at the University of Alabama by Shyi Ming Chen and Chia-Ching Hsu in 2004 [6]. The results of the comparison of the fuzzy time series method proposed by them can get a higher forecasting accuracy rate for forecasting enrollments than the existing methods [7]. A new method to forecast enrollments using fuzzy time series and clustering techniques [8]. The experimental results show that the proposed method gets a higher average forecasting accuracy rate than the existing methods. Furthermore, the process of this method does not require a learning system from a complex system, as it does in genetic algorithms and neural networks so the method is easy to develop [9].In this research, the forecasting application of national motorcycle sales using method of high order fuzzy time series combined with interval determination using automatic clustering algorithm was

Contrasting Univariate and Multivariate Time Series Forecasting Methods for Sales: A Comparative Analysis

“…Time series algorithms are commonly used for forecasting sales volume, primarily relying on the seasonal fluctuations of sales volume (ramosal et al, 2022). To improve the accuracy of forecasting, time series algorithms can be combined with other algorithms (Rasim et al, 2018). The growing popularity of e-commerce has enabled the use of computer technology for sales volume forecasting, although it is essential to consider multiple influencing factors.…”

Achieving Sales Forecasting with Higher Accuracy and Efficiency: A New Model Based on Modified Transformer

“…The historical datasets typically exhibit irregular variations due to a myriad of real-world factors such as natural disasters, seasonal fluctuations, adaptations in marketing strategies, random influences, and unforeseen contingencies. Large-scale sales datasets in real-world scenarios possess a complexity characterized by nonlinearity, diversity, persistence, and noise [19]. These datasets often contain significant gaps with missing attribute values or are limited to aggregated data.…”

Prediksi Tingkat Penjualan Sepeda Motor dengan Metode Rough Set

“…Pada penelitian lain tahun 2020, metode rough set juga digunakan untuk melakukan prediksi penjualan perumahan, dimana atribut yang digunakan adalah pekerjaan, tipe rumah, kemampuan ekonomi dan bentuk pembayaran, sehingga menghasilkan 5 reduct dan 11 rules [14]. Pada tahun 2018, penelitian tentang peramalan penjualan sepeda motor pernah dilakukan dengan metode clustering dan Fuzzy Time Series, penelitian ini menggunakan data penjualan sepda motor Suzuki selama 10 tahun dari tahun 2005 sampai 2014 dengan hasil penelitian menunjukkan nilai rata-rata MPE sebesar 0,19% dan MAPE sebesar 2,15% termasuk dalam akurasi yang baik [15]. Penelitian sebelumnya tentang peramalan panjualan sepeda motor pada tahun 2020, peramalan penjualan sepeda motor dengan merek Yamaha di Sentral Yamaha Malang menggunakan data 3 tahun penjualan.…”