DESIGN OF GOLD, SILVER, DOLLAR PRICE PREDICTION APPLICATIONS USING WEB-BASED REGRESSION ALGORITHM

Design of future gold, silver, and dollar price prediction applications using linear regression, ridge regression, and web-based lasso algorithms

Abstract

To provide services and help people who want to invest in gold, silver, and dollars, this paper discusses the creation of applications which can be use by public in accessing information about the price movements of gold, silver, and dollars against rupiah. Based on the knowledge of algorithms, linear regression, ridge regression and lasso algorithms, application for predicting the price of gold, silver, dollars for tomorrow were created, using the flask web application. In making this application, it utilizes a database management system PostgreSQL, as gold, silver, and dollar dataset storage. The dataset is obtained from sources that provide daily price movements of gold, silver, and dollar. The testing process starts with the user opening linked share. After entering the application page, user selects type to be predicted, after that chooses algorithm method to be used. Then application will display prediction results for next day.