The main goal of this project is to combine the power of machine learning algorithms, especially multiple regression, regressor tree decision and Random forest, to predict phosphate prices. By incorporating historical data that includes many important factors, such as diesel price, phosphate and diesel ratio, diesel ROC price and more, the project aims to develop robust forecast models.
This effort includes careful data processing, including handling missing values and outliers, and thoughtful feature selection. The implementation of these algorithms allows valuable insights to be extracted from the data, thus enabling accurate predictions of future phosphate prices. With robust evaluation metrics such as absolute error, root mean square error, root mean square error and R-squared, the model performance will be carefully evaluated, ultimately identifying the model with the strongest predictive ability. In addition of vizulaization charts and graphs with POWER BI.
The project aims to provide phosphate industry managers with reliable tools for making smart decisions in the dynamic field of phosphate.
