ANALISIS PERBANDINGAN METODE REGRESI LINIER DAN NON-LINIER PADA SAMBUNGAN MATERIAL PELAT BAJA KARBON ASTM A36

The joining of ASTM A36 carbon steel plates using the Shield Metal Arc Welding (SMAW) technique is influenced by the tensile strength and hardness of the joint. To improve the quality of ASTM A36 carbon steel plate joints in the future, prediction methods are used. The prediction results play a significant role in determining the quality of the material connection.The method used in this research was descriptive quantitative, by predicting linear regression and non-linear (quadratic) regression models. The purpose of the discussion in this article was to determine the best prediction model for current variations on the tensile strength of future material connections. The analysis of the correlation coefficient showed a strong relationship in both models discussed. This indicates that the current strength variable had a strong relationship with tensile strength. Meanwhile, the highest R-square score was obtained in the non-linear regression model. The determination of the best prediction method was found in the non-linear regression model on tensile strength, with a MAPE value of 3.57%. These results indicate that the non-linear regression model better describes the relationship between current strength and tensile strength in ASTM A 36 carbon steel plate joining.