Time series modeling analysis on halal certificate issuance patterns for msme products at the halal center of Airlangga University

The halal industry has become one of the strategic sectors in Indonesia's economic development, especially with the increasing public awareness of the importance of consuming halal products. In line with the mandate of Law No. 33 of 2014 concerning Halal Product Assurance, all products entering, circulating, and traded in the territory of Indonesia must be halal certified. The Halal Center of Airlangga University is one of the institutions assisting the halal certification process that actively serves Micro, Small, and Medium Enterprises (MSMEs) in issuing halal certificates. However, the process of issuing halal certificates does not always take place consistently every period. Fluctuations in the number of applications and the issuance of certificates can be influenced by various factors. Through time series modeling analysis, the pattern of halal certificate issuance can be mapped more systematically. By modeling historical data on the number of certificates issued, trends, seasonality, and fluctuations that occur can be identified, as well as producing predictions for future periods. This information is important for the Halal Center of Airlangga University in designing more efficient resource planning and service strategies. The research began with a more in-depth data analysis with Exploratory Data Analysis to identify data characteristics. Then, enter the modeling process that begins with data splitting with a scale of 85:15, and continues by dividing the training data and testing data. The models used for time series modeling analysis are the ARIMA, LSTM, and Hybrid (ARIMA-LSTM) models. Evaluation of the three model algorithms shows that the LSTM model is superior, with an MAPE evaluation of 10.075% with an MAE of 12, and an RMSE value of 31.279. These results explain that the LSTM model is the most optimal model for forecasting halal certification issuance patterns at the Halal Center of Airlangga University, so that forecasting results are obtained that show experience issuance patterns that tend to increase significantly Keywords: Time Series Modeling, Publishing Pattern, Halal Certification, Halal Center, ARIMA, LSTM, Hybrid (ARIMA-LSTM)