Journal of Computer Science and Information Technology

[Link]
2023 Volume 9 Is s u e 2 Pg 72-76 e-ISSN: 2502-1486

Prediction of Sales and Inventory using the EOQ Method based on

Single Exponential Smoothing Forecasting

Alfin Andika Putra* , Vicky Ariandi, Pradani Ayu Widya Purnama

Universitas Putra Indonesia YPTK Padang, Jl. Raya Lubuk Begalung Padang, West Sumatra – 25221, Indonesia
*alfinandika93@[Link]

Abstract
Along with the development of the era where technology is becoming more sophisticated like today, the need is also
increasing. Science and technology have also experienced drastic progress. Forecasting is a method for making estimates of
future data involving the use of past data in the form of a systematic model, and the forecasting method used is Single
Exponential Smoothing. Planning the Economic Order Quantity (EOQ) method can minimize the occurrence of out of stock
so that there is no disruption to processes in a business and is able to save costs due to the efficiency of inventory at the
business premises concerned. Implementation of Supply Chain Management (SCM) with a data mining system at CV. Amifa
Keluarga Lestari can simplify the management of raw materials by implementing an SCM system which can reduce excess
stock purchases. With the Single Exponential Smoothing method, you can predict the number of best sales for the next month
in one period by looking at the smallest error. The calculation results show that the most economical order in one order is
1291 Kilograms, and the Total Storage Cost is Rp. 154,919 per Kilogram. The use of EOQ provides a number of benefits in
company inventory and sales management, increasing efficiency and reducing risk. It also optimizes operational efficiency
by saving time and resources in inventory management.

Keywords: Supply Chain Management, Single Exponential Smoothing, EOQ Method, Forecasting, Inventory

JCSITech is licensed under a Creative Commons 4.0 International License.

The integration of activities that procure materials and
1. Introduction
services, convert them into intermediate goods and
Along with the development of the era where final products, and deliver them to customers is referred
technology is becoming more sophisticated like today, to as supply chain management. This concept
the need is also increasing. Science and technology prioritizes the transformation of raw goods into finished
have also experienced drastic progress. An example of goods that are sent to customers [3].
technology that is progressing is laptops. Nowadays
Forecasting is a method for making estimates of future
laptops have become a basic need for people to do
data by involving the use of past data in the form of a
business, education, play games and even create games,
systematic model. The forecasting model used in
communicate and so on. Not only that, laptops can also
forecasting cracker sales is the Time Series or what is
be taken anywhere so that these activities can be done
usually called a time series and the forecasting method
at any time and more easily [1]
used is Single Exponential Smoothing. The reason the
All businesses basically plan and control goods with author chose this method is because compared to
the main aim of minimizing costs and maximizing several other forecasting methods such as Single
profits. In planning and controlling goods, the problem Moving Average and Double Exponential Smoothing,
is how to organize the most appropriate inventory of the Single Exponential Smoothing method provides
goods so as not to disrupt production activities and the better forecasting accuracy because it produces fewer
costs required are not excessive [2] errors if the data used has a stationary data pattern [4]
Current global competition has given birth to The single exponential smoothing method is a
companies that have succeeded in adapting and development of the single moving averages method
innovating. This success is also influenced by the where this forecasting method is carried out by
ability to utilize technology , communication and repeating calculations continuously using the latest data
information. The global market requires every company and each data is given a weight [5]. Single Exponential
to compete, with increasingly high expectations for Smoothing uses very little past data recording. This
products and services, and increasingly shorter product model assumes that data fluctuates around a fixed
life cycles, has encouraged companies to prioritize their average value, without following a pattern or trend [6].
focus in their supply chains to achieve competitive
advantages that can support direct business processes .

Accepted: 02-19-2022 | Revised: 03-16-2023 | Published: 04-30-2023 | doi: 10.35134/jcsitech.v9i2.66

1
Alfin Andika Putra, et al
With an inventory policy implemented in the company,
inventory costs can be kept as small as possible. To
overcome this problem, a management strategy is 𝑆𝑡 = 𝑎 ∗ 𝑋 − 1 + (1 − 𝑎) ∗ 𝑆𝑡 − 1
needed in the form of forecasting sales and inventory in
preparation for meeting customer demand in the
coming period . EOQ method planning. This can Where,
minimize the occurrence of out of stock so that St+1 = Forecasting Results
processes in a business are not disrupted and can save α = Constant value ( 0<α<1)
costs due to efficient inventory of goods at the business Xt = Actual Value/quantity of goods
premises concerned [7]. St = Experience on the previous day
EOQ is the value of the amount of materials needed
during each purchase using the most economical costs The overall accuracy of any forecasting model be it
[8]. EOQ (Economic Order Quantity) is the number of moving average, exponential smoothing or others can
goods that can be purchased at minimal cost. The aim be explained by comparing the projected values with
of this method is to determine the number of orders that the actual or observed values. The level of forecasting
can minimize storage costs and inventory ordering accuracy can be measured from the following values:
costs [9]. Previous research using the EOQ method is
widely used, such as raw material inventory [10] 1. Mean Absolute Error (MAE)
[11][12], product sales analysis[13], generic drug Mean Absolute Error (MAE ) is one of the methods
inventory[14], and others. used to measure the level of accuracy of a forecasting
model. The MAE value shows the average absolute
2. Research methodology error between the forecasting/prediction results and the
real value. The MAE formula is explained as follows:
To assist in preparing this research so that the steps in
solving the problem to be discussed can be arranged 𝑛
1
clearly, a framework structure is needed. The research 𝑀𝐴𝐸 = ∑ |𝑓𝑖 − 𝑦𝑖 |
framework is in Figure 1. 𝑛 𝑖=1

Where,
𝑀𝐴𝐸= Mean Absolute Error
𝑓𝑖 = value of forecasting results,
𝑦𝑖 = true value,
𝑛 = amount of data

2. Mean Squared Error (MSE).

Mean Squared Error (MSE) Is the average sum of
squared forecasting errors. By using MSE, the error
shows how much the forecasting results differ from the
results to be predicted. The formula equation used is
Mean Squared Error (MSE).

1
𝑀𝑆𝐸 = ∑ |𝑋𝑡 − 𝑆𝑡|²
𝑛
Information:
MSE = Mean Square Error
Xt = Actual value for the t-th period
Figure 1. Research Framework
St = Forecast value for the t-th period
In this method, forecasting is carried out by repeating
3. MAPE(Mean Absolute Percentage Error)
calculations continuously using the latest data, where
Mean Absolute Percentage Error (MAPE) is a
each data is given a weight. The α value is the weight
calculation using the absolute error in each period
value in the Single Exponential Smoothing method.
divided by the real observation value for that period.
The α value is used to refine the forecasting method.
Then, average the absolute percentage errors. The
The amount of α is determined using the trial and error
smaller the MAPE value, the better the level of
method. Where the determination is calculated
prediction accuracy. The formula equation used is
sequentially with the condition 0<α<1. The α value to
Mean Absolute Percentage Error (MAPE).
be tried is 0.1 to 0.9 . In this method, an alpha value has
been determined to assist the calculation process in
order to obtain appropriate forecasting results . The
|𝑋𝑡 − 𝑆𝑡|
equation of Single Exponential Smoothing is as ∑
𝑀𝐴𝑃𝐸 = 𝑋𝑡 𝑥 100%
follows:
𝑛

Journal of Computer Science and Information Technology Volume. 9 Issue 2 (2023) 72-76
2
Alfin Andika Putra, et al
Where 𝑆𝑡6 = 0.1 ∗ 1614 + (1 − 0.1) ∗ 4665.2 = 4360.1
Xt = actual value on day t
St= forecasting value on day t Forecasting search with alpha value = 0.2

4. Economic Order Quantity (EOQ) Table 2. Forecasting data alpha 0.2

Forecast Mean Absolute Mean Squared
No MAPE
(St) Error Error
EOQ is the number or size of orders
owned , the minimum amount of ordering costs and
1 5166 0 0 0
carrying costs per year. After the minimum number of 2 5166 1152 1327104 28.6995517
materials purchased is determined, the next problem 3 4935.6 820.6 673384.36 19.9416769
that arises is when the company should reorder so that 4 4771.48 2734.48 7477380.87 134.240548
the company does not run out of materials. The formula 5 4224,584 2610.54 6815148.81 161.7462206
6 3702.4672
equation used by Economic Order Quantity (EOQ) is as MEANS 522 272588.41 162
follows:
Table 3. Forecasting Results Data for June 2022
No Forecast MAE MSE MAPEAlpha(α)
2𝑆𝐷 1 4360.7 305 93098,214 189% 0.1
𝐸𝑂𝑄 = √ 2 3702.6 522 272588.41 162% 0.2
𝐻
3 3710.4 898 807183.6523 186% 0.3
Where, 4 2743.6 753 566870.4564 117% 0.4
EOQ = Economic order quantity 5 2404.7 790 624692.6406 98% 0.5
D = expected demand per time period 6 2138.7 787 619464.7028 81% 0.6
7 1934,2 748 559339.1529 66% 0.7
S = Order Order Cost 8 1782.5 675 455646.6002 52% 0.8
9 1677.94 568 322449.07 39% 0.9
3. Results and Discussion
From the table above you can see the results of the
This research uses data on sales of [Link] Keluarga sales data forecasting calculations. So the results of
Lestari crackers in the period January 2022 to May forecasting using the Single Exponential Smoothing
2022. method in June 2022 with the smallest error percentage
Table 1. Total Sales Data for 2022
value are α = 0.9 with Forecast results of 1677.094 and
Sales MAPE 39.00%.
No Month Actual
Turnover
January 92,990,400 5166
3.2 Economic Order Quantity
1
February 72,259,760 4014
2
3
March 74,073,280 4115 Stages for calculating EOQ and TIC to find out the best
April 36,681,440 2037 total order and efficient costs.
4
May 29,060,080 1614
5
1. Calculating EOQ
3.1 Single Exponential Smoothing 2. Calculating TIC
Based on data from the object, it is known:

In this method, an alpha value has been determined to Table 4. Stock Data
assist the calculation process in order to obtain Number of Goods Required (D) 10,000 Kg/month
appropriate forecasting results Order Fee (S) Rp. 10,000/Kg
Storage Costs (H) Rp. 120/Kg

𝑆𝑡 = 𝑎 ∗ 𝑋 − 1 + (1 − 𝑎) ∗ 𝑆𝑡 − 1
Based on the table above we can calculate EOQ and
TIC with the following formula:
Forecasting search with alpha value = 0.1

St1=5166 (the first forecast data still cannot be 𝐸𝑂𝑄 = √2(𝐷. 𝑆) ∕ (𝐻)
calculated because there is no forecast data for the
previous month). 𝐸𝑂𝑄 = √2(10000 × 10000)⁄(120)
200.000.000
=√ = √1.666.666 = 1291
𝑆𝑡2 = 0.1 ∗ 5166 + (1 − 0.1) ∗ 5166 = 5166 120

𝑆𝑡3 = 0.1 ∗ 4014 + (1 − 0.1) ∗ 5166 = 5050.8 𝐷 𝑄

𝑇𝐼𝐶 = ( × 𝑆) + ( × 𝐻)
𝑄 2
𝑆𝑡4 = 0.1 ∗ 4115 + (1 − 0.1) ∗ 5050.8 = 4957.22
10000 1291
𝑇𝐼𝐶 = ( × 10000) + ( × 120)
𝑆𝑡5 = 0.1 ∗ 2037 + (1 − 0.1) ∗ 4957.22 = 4665.2 1291 2

Journal of Computer Science and Information Technology Volume. 9 Issue 2 (2023) 72-76
3
Alfin Andika Putra, et al
100.000.000 4. Forecasting Calculation Page
=( 1291 ) + (77.460)
=(77.459) + (77.460) = 154.919 The image below is the admin page for adding
forecasting data to predict next month's sales.
So, it can be seen that the most economical order in one
order is 1291 Kilograms, and the Total Storage Cost is
Rp. 154,919 per Kilogram.

3.3 System Testing

1. Raw Material Data Page

On the data page raw materials where admin users can

add, edit and delete raw material data in Figure 2
below.
Figure 5. Calculate Forecasting Page

5. Forecasting Results Page

The image below is a page for calculating sales

forecasting to predict sales for the following month.

Figure 2. Raw Material Data Page

2. Purchase Transaction Page

On the purchase transaction page where admin users

can add, edit and delete purchase data in Figure 3 Figure 6. Forecasting Results Page
below.
6. Print Page

On this page the owner can print a report on the results

of purchase transactions and forecasting results that
have been calculated.

Figure 3. Purchase Transaction Page

3. Sales Transaction Data Page

The image below is the sales transaction page where

admin users can add, edit and delete sales data. In
Figure 4 below Figure 7. Sales Transaction Print Page

Figure 4. Sales Transaction Data Page

Journal of Computer Science and Information Technology Volume. 9 Issue 2 (2023) 72-76
4
Alfin Andika Putra, et al

4. Conclusion [12] Dewi, PCP, Herawati, NT, & Wahyuni, MA (2019). Analysis
of Inventory Control Using the Economic Order Quantity
Based on direct research and analysis carried out at CV. (EOQ) Method to Optimize the Supply of Mineral Water
Amifa Keluarga Lestari can conclude that the Packaging Raw Materials (Case Study at PT. Tirta Mumbul
Jaya Abadi). Journal of Professional Accounting, 10(2), 54-65.
implementation of Supply Chain Management with a [Link]
data mining system can make it easier to manage raw [13] Nugraha, YA, Niswatin, RK, & Helilintar, R. (2022).
materials by implementing an SCM system which can Application of the EOQ Method for Analysis of MSME
reduce excess stock purchases which can cause excess Product Sales in Nganjuk Regency. Generation Journal, 6(1),
13-21. [Link]
stock so that expenditure costs increase. The system is [14] Dyatmika, SB (2018). Generic Drug Inventory Control Using
equipped with a sales transaction menu for each month the ABC Analysis Method, Economic Order Quantity (EOQ)
so that predictions will be calculated from the previous Method, and Reorder Point (ROP) at XYZ Pharmacy in 2017.
month's sales. And with the Single Exponential Modus, 30(1), 87-95.
[Link]
Smoothing method , you can predict the number of best
sales for the next month in one period by looking at the
smallest error .
References
[1] Primawanti, EP, & Ali, H. (2022). The Influence of
Information Technology, Web-Based Information Systems and
Knowledge Management on Employee Performance
(Literature Review Executive Support Systems (ESS) for
Business). Journal of Information Systems Management
Economics, 3(3), 267-285
[Link]
[2] Rusida, ND, & Noer, ZM (2018). Design of software to help a
desktop application-based sales system at the Pangandaran
Instamie Cafe. Journal of Information Management and
Engineering (JUMANTAKA), 1(1).
[3] Yun, Y., & Kurniawan, A. (2017). Supply Chain Management
Analysis of Competitive Advantage in Food Production
Cooperatives in West Bandung Regency. GEMA: Journal of
Gentiaras Management and Accounting, 9(2), 160-172.
[4] Gustriansyah, R. (2017). Analysis of the single exponential
smoothing method with brown exponential smoothing in a
case study predicting the quantity of sales of pharmaceutical
products in pharmacies. Semnasteknomedia online, 5(1), 3-5.
[5] Ginantra, NLWSR, & Anandita, IBG (2019). Application of
the Single Exponential Smoothing Method in Forecasting
Sales of Goods. J-SAKTI (Journal of Computer Science and
Informatics), 3(2), 433-441. [Link]
sakti.v3i2.162
[6] Pakpahan, HS, Basani, Y., & Hariani, RR (2020). Prediction of
the Number of Poor Population in East Kalimantan Using
Single and Double Exponential Smoothing. Inform.
Mulawarman J. Ilm. Computing Science, 15(1), 47-51.
[Link]
[7] Hidayat, K., Efendi, J., & Faridz, R. (2020). Analysis of raw
material inventory control for raw potato crackers and curly
potatoes using the economic order quantity (EOQ) method.
Performance: Industrial Engineering Scientific Media, 18(2).
[Link] .
[8] Umami, DM, Mu'tamar, MFF, & Rakhmawati, R. (2018).
Inventory Cost Efficiency Analysis Using the Eoq (Economic
Order Quantity) Method at Pt. Xyz. Journal of
Agrotechnology, 12(01), 64-70. [Link]
agt.v12i1.8100
[9] Rachmawati, SA, Syafirullah, L., & Faiz, MN (2020,
November). Design of an Inventory Control System Using
Web-Based EOQ and ROP Methods. In Proceedings of the
National Seminar on Applied Innovative Research
(SENTRINOV) (Vol. 6, No. 1, pp. 778-786).
[10] Mayasari, D. (2021). Analysis of Raw Material Inventory
Control Using the EOQ (Economic Order Quantity) Method at
PT. Suryamas Lestari Prima. Bis-a, 10(02), 44-50.
[Link]
[11] Ismaya, YB, & Suseno, S. (2022). Analysis of Sweet Potato
Raw Material Control Using Economic Order Quantity (EOQ)
and H-Sin Rau PT Methods. Galih Aesthetic Indonesia.
Journal of Applied Industrial Technology and Management,
1(2), 123-130. [Link]

Journal of Computer Science and Information Technology Volume. 9 Issue 2 (2023) 72-76
5