Prediksi dan Analisis Time Series pada Data COVID-19

Abstract

Data processing on a large scale is currently needed for various needs in the field of information technology to create conclusions that are easy to understand and analyze based on existing data. With methods in data science, large-scale data processing makes it easier to present data to be understood and analyzed. This research was conducted by processing COVID-19 data based on the Time Series. The steps in this research are to apply Exploratory Data Analysis first, then visualize the data, and make predictions using the ARIMA and Prophet methods. The research was conducted with the aim of processing COVID-19 data into informative and easy-to-analyze data, visualizing the COVID-19 dataset to make it easier to understand, and making predictions for the future. The dataset used was obtained from Kaggle entitled "COVID-19 data from John Hopkins University" which contains confirmed data and death data from around the world. From the dataset, several countries in Southeast Asia were selected for exploration, including Indonesia, the closest country to Indonesia, and Southeast Asian countries. From the exploration results obtained various information from data in the form of a DataFrame which is easy to analyze, a variety of graphic plots that are easy to understand, and get the prediction results of confirmed cases in Indonesia from the ARIMA and Prophet methods which are then determined that the optimal prediction is using the Prophet.