Freight Cost Prediction Using Machine Learning Algorithms

With the advent of the digital age, it is certainly noticeable that the world has indeed become a smaller place and better connected due to progressing technology and innovations that occur every day. In the virtual world, the distance factor has certainly been negated due to the Internet phenomenon, however in the physical world, distance plays an extremely important role in any kind of business decisions that a company might need to take, and the cost related to the distance is also one of the most prominent factors that influence any decision. This paper focuses on predicting the freight cost of a cargo, which is dependent on a variety of variables such as location, the weight of the cargo, etc. Four widely applied ML models are adopted, viz. K nearest neighbors regressor, Random Forest regressor, XGBoost regressor, and LightGBM regressor. Using the most prominent and highly influential features out of all, the comparative analysis of the machine learning models is presented.