<![CDATA[The main objective of this research is to design a web-based machine learning model that can predict the mechanical properties of aluminum based on its chemical composition. By inputting nine variables of chemical elements such as Al, Mg, Zn, Ti, Cu, Mn, Cr, Fe, and Si, the model is able to provide predictions for two output data, Yield Strength (YS) and Tensile Strength (TS). The research aims to understand the relationship between chemical composition and mechanical properties of aluminum, and to develop a tool that can be used to predict these properties with a high level of accuracy. Overall, the goal of this study is to enhance the understanding of the properties of aluminum and how it can be utilized in various applications. This study designs a web-based machine learning modeling to predict the mechanical properties of aluminum in the percentage of chemical composition, where the input data in the modeling consists of 9 variables of chemical elements such as Al, Mg, Zn, Ti, Cu, Mn, Cr, Fe, Si, and has 2 output data consisting of Yield Strength (YS) and Tensile Strength (TS). The modeling machine learning is designed using the Python programming language and additional libraries such as Pandas, Numpy, Scikit-learn, and Streamlit. The modeling in this study uses three algorithms consisting of Decision Trees (DT), Random Forest (RF), and Artificial Neural Network (ANN). Each algorithm is optimized with the best search parameters, and where the RF algorithm has better performance than DT and JST. The best modeling uses the RF algorithm with optimal parameters of number of trees at 20 and maximum depth of 10, with MAE values of 11.44, RMSE of 14.282, and R of 0.93 for Yield Strength (YS) predictions, and for Tensile Strength (TS) predictions, MAE values are obtained. 21,669, RMSE 27,301, and R 0.871. ]]>
Copyrights © 2023
