<![CDATA[Employee attrition is a serious problem that must be considered by every company, including PT Kimia Farma Tbk. High employee turnover rates can affect company productivity and performance, negatively impacting the business. This study aims to analyze the effect of Feature Engineering on the Logistic Regression and Random Forest methods on the prediction of employee attrition at PT Kimia Farma Tbk. In addition to knowing the most effective method in increasing employee attrition prediction at PT Kimia Farma Tbk. The results of this study indicate that feature engineering significantly affects performance in predicting employee attrition at PT Kimia Farma Tbk. using Logistic Regression and Random Forest models. It can be seen that the application of feature engineering can affect the accuracy, precision, recall, and F-Score of the two methods. The Recursive Feature Elimination (RFE) method with the Logistic Regression model has an accuracy of 0.866, a precision of 0.5, a recall of 0.159, and an F-Score of 0.259. Meanwhile, the RFE with the Random Forest model has an accuracy of 0.886, a precision of 0.916, a recall of 0.25, and an F-Score of 0.392. The SelectKBest method with the Logistic Regression model has an accuracy of 0.88, a precision of 0.9, a recall of 0.204, and an F-Score of 0.333. Meanwhile, SelectKBest with the Random Forest model has an accuracy of 0.87, a precision of 0.818, a recall of 0.204, and an F-Score of 0.327. According to the results of the performance comparison, the RFE (Recursive Feature Elimination) method with the Random Forest model can be said to be the best method in terms of accuracy and precision. Although the recall of this method is slightly lower, the performance of this method still meets the criteria as a good method. Therefore, the Recursive Feature Elimination method with the Random Forest model was chosen as the best method for this case.]]>
