Linear Regression Method To Determine Attributes That Affect Student Achievement

Abstract

Education is basically to produce generations or individuals who excel in thinking, acting and making decisions. The measure used in assessing the character of a student at school is grades. To get a good score is certainly influenced by several factors. This study tries to analyse any factors with linear regression method that affect student success in terms of achievement. The method used is 1. Converting data to numeric 2. Deleting missing values ??3. Eliminating duplicate data 4. Normalizing data (range transformation) 5. Selecting the target attribute and then looking for correlation with other attributes with the highest positive value . 6. Separate the data into testing data and training data with a ratio of 9:1. 7. Look for the performance of the model made. 8. Change parameters to get the highest performance. The results of the calculations and the built model show that the attributes fedu, medu, g1 and g2 are the determinants in improving student achievement. The G2 attribute has the greatest influence on student performance in learning.