IMPLEMENTATION OF SINGLE EXPONENTIAL SMOOTHING METHOD DEMAND FOR HERBAL MEDICINE TO DC RIA SARI ANGGRIANI

: Distributor Center (DC) Ria Sari, one of the official herbal medicine agents from PT HPAI (Herba Penawar Alwahida Indonesia) company. Sales of herbal medicines at the Ria Sari Anggriani Distributor Center (DC) often experience instability. Where in the period March 2022 to May 2022 there was a significant increase in sales. This resulted in the Ria Sari Anggriani Distributor Center (DC) being unable to serve requests for available herbal medicines. So that it becomes an obstacle in the stock of herbal medicines that must be prepared. If there are too many drug stocks but the demand for herbal medicines is decreasing then this will result in a loss, conversely if the demand for medicines increases but the stock of herbal medicines cannot be prepared then this will be a loss for DC Ria Sari Anggriani. For this reason, it is necessary to do a forecasting/prediction technique, using the Single Exponential Smoothing Method . The purpose of the forecasting/prediction will be to assist in terms of the supply of herbal medicines in the following month's period. The results of testing this method were obtained in the period June 2023, requests for 167 herbal medicines, 178 Extra foods, 189 Etta Goat Milk, 91 Herbal Toothpastes, 74 products of Propolis Soap.


INTRODUCTION
Herbal medicines come from plants that are processed or extracted into powders, pills or liquids without the use of any chemicals.As is known, herbal medicine can cure diseases with minimal side effects because it is different from synthetic drugs, herbal medicines are made from natural ingredients which can cause side effects immediately and in the long term [1].The practice of using herbal medicines is widespread in many countries in the world, especially in countries with biodiversity such as Indonesia which has great potential in the development of traditional medicines [2].The use of herbal medicines as traditional medicines has been widely consumed by the public at this time, and obtaining these drugs is also not difficult.
Distributor Center (DC) Ria Sari Anggriani is one of the agents or distributors for selling herbal medicines from the company PT HPAI ( Herba Penawar Alwahida Indonesia) having its address in Hamlet I Desa Subur, Kec.Joman Water.This business has been in business for about 5 years.The herbal medicine business products include synergy herbal oil, extrafood , etta goat milk, herbal toothpaste and propolis soap.
In the process of selling herbal medicines at DC Ria Sari Anggriani a problem arose, the results of the discussions that were carried out were the instability of sales of herbal medicinal products every month.From 2022 sales data In early March to May the demand for herbal medicines suddenly increased.This can result in a decrease in service quality so that DC Ria Sari Anggriani can lose the trust and loyalty of its customers.Due to limited drug stock, customers are forced to look for other stores that sell the same product.If this continues to happen, DC Ria Sari Anggriani's custom-ers will continue to decrease over time.Then in another condition, the demand for herbal medicines is at a low point, but DC Ria Sari Anggriani has provided as much product stock as possible, so that the existing supplies will not be sold out.This can be a loss for DC Ria Sari Anggriani because herbal medicines also have an expiration date like other medicines, and if that happens, expired medicines cannot be resold.Not to mention the cost of maintaining and storing herbal medicines to keep them in good condition and fit for use .
For this reason, measurement is needed as a solution to predict herbal medicine stocks and help DC Ria Sari Anggriani to overcome problems in terms of maintaining stability of herbal medicine stocks using forecasting techniques.The purpose of this forecasting measurement is later to find out how far the level of accuracy generated by forecasting herbal medicine stocks is so that the accuracy of forecasting herbal medicine stocks is known in the next month.
One technique that can be used to predict a value or event in the future is forecasting using past data or information as a reference for planning and decision making [3].Accurate forecasting is one of the essential elements for effective management.The purpose of forecasting is to reduce uncertainty and to provide benchmarks (benchmark) to monitor actual performance [4].The goal of forecasting is to get predictions that can minimize forecast errors which are usually measured by Mean Absolute Deviation (MAD) and Mean Squre Error (MSE) [5].
The method used in the forecasting used in this study uses the Single Exponential Smoothing method .Single exponential smoothing method used be-cause the technique used to record past data is very small and assumes data that fluctuates or is not settled [6].
Forecast accuracy is something that is very important in forecasting because it measures the suitability of existing data with forecasting data.In this study, researchers performed calculations of predictive value determination by calculating the mean squared error value (Mean Squared Error (MSE), and the absolute percent error.average (Mean Absolute Percentage Error (MAPE) [7].The purpose of this study is to find out how far the level of accuracy generated by forecasting on DC Ria Sari Anggriani's herbal medicine stock is known so that the accuracy of forecasting herbal medicine stocks in the next month is known based on past sales data.
Much research on forecasting have been carried out, among others.In Handokos research (2019) the single exponential smoothing method can be used to predict the number of new student admissions so that the predicted value can be known in the next period [8].According to Nangi (2018), the results of his research are predictive applications of drug stock data in the pharmaceutical installations of RSUD Kab.MUA was successfully constructed by applying the triple exponential smoothing (TES) method [9].According to Fahrunnisa (2021) in his research applying the Single Exponential Smoothing method was able to provide predictions of hypertension in the Rawang Panca Arga Health Center the following month [10].In research Fachrurrazi (2015) also explained that by applying the Single Exponential Smoothing method it can be used to predict the amount of drug inventory, in predicting sales of the Ambeven drug in March 2015 there were 49 tablets, at the Bintang Geuruguk drug store [11].Indah (2018) also explained in her research that SES is a single smoothing method with parameters alpha 0.2 and alpha 0.4, Single Exponential Smoothing makes comparisons in determining alpha to produce the smallest forecasting error and the minimum forecasting error that will be selected for look for forecasts for the next period [5].

METHOD
Methodology is a stage or theoretical framework intended to carry out research in certain ways to obtain truth.This research methodology applies quantitative research, namely an approach in research that uses data in the form of numbers or quantitative data to collect information, analyze phenomena, and look for patterns and relationships between the variables studied.
This approach focuses on collecting drug sales data through statistical methods and mathematical analysis to generate conclusions or generalizations.The stages in this study contain a research framework with several stages including problem identification, problem analysis, goal setting, literature study, what data is collected, data analysis, analysis of the SES and UMK methods, model design, system implementation, testing of results and results and discussion.The single exponential smoothing method uses the formula equation in equation ( 1): Where: Ft+1 = Forecast for period t+1 Xt = real value of t period α = Weight indicating smoothing constant (0 < α < 1) Ft-1 = Forecast for period t-1 Furthermore, to measure the level of error, it can be tested using MSE, the smaller the Mean Squared Error (MSE) value, the more accurate the prediction.

RESULTS AND DISCUSSION
The first step is to analyze the system.Historical data on demand for herbal medicines will be calculated using the Single Exponential Smoothing method to look for predictions of the number of requests for herbal medicines in the following period/month.Then the predicted value will be tested or the error value calculated using the Mean Squared Error (MSE) method.so that in the end the prediction value with the smallest error will be obtained.As for the data that will be used in this study, historical data on the number of requests for herbal medicines at DC Ria Sari Anggriani, from January 2022 to May 2023.The data on demand for herbal medicines can be seen in table 1

System analysis
Based on historical data on de-mand for herbal medicines from January 2022 to May 2023, researchers will look for predictions of demand for herbal medicines in the upcoming June 2023 period.The predicted value will be searched using the Single Exponential Smoothing method with an alpha value of 0.1 to 0.9.Then because the value of a prediction still contains errors.Then the error value will be searched using the Mean Squared Error method for each alpha value.

Synergy Herbal Oil Demand Forecasting
In the calculation of synergy herbal oil forecasting the alpha value used is 0.9 for the number of requests data because of all the alpha values that have been calculated that the value α = 0.9 has a small error .Ft +18 = α Y 17 +(1-α) F 17 = (0.9*168)+(1-0.9)167.25 = 151.2+16.725= 167.925

Demand Forecasting Extrafood
Extrafood forecasting calculation the alpha value used is 0.1 for the number of requests data because of all the alpha values that have been calculated that the value α = 0, 1 has a small error .

Etta Goat Milk Demand Forecasting
Etta Goat Milk forecasting calculation the alpha value used is 0.9 for the number of requests data because of all the alpha values that have been calculated that the value α = 0.9 has a small error .

Forecasting Herbal Toothpaste Demand
In the forecasting calculation of Herbal Toothpaste the alpha value used is 0.6 for the number of requests data because of all the alpha values that have been calculated that the value α = 0.6 has a small error .

Forecasting the Demand for Propolis Soap
In forecasting calculations for Propolis Soap the alpha value used is 0.9 for the number of requests data because of all the alpha values that have been calculated that the value α = 0.9 has a small error .Anggriani, so that the predicted demand value can be known in the following month's period.As well as the Mean Squared Error (MSE) method can be used to find the error value (error) in each alpha value so that the smallest error value will be obtained.

Table 1 .
. Herbal Medicine Demand Data in January 2022 -May 2023

Table 2 .
Synergy Herbal Oil Demand Forecasting Analysis

Table 4 .
Etta Goat Milk Demand Forecasting

Table 5 .
Herbal Toothpaste Demand Forecasting

Table 6
The results of the calculation and the value of the Error Forecasting Forecasting Demand for Herbal medicines at DC Ria Sari Anggriani can be seen in the table7.Table7.Forecasting Results and Forecasting Error Values 167.92 for Herbal Oils.Forecasting calculations that have been carried out and based on the Error Forecastin results table, several conclusions can be drawn, namely the Single Exponential Smoothing Method can be used to predict the number of requests for herbal medicines at DS Ria Sari is