Prediction of Downstream BOD based on Light Gradient Boosting Machine Method

Abstract

The problem of water pollution has been one of the most concerned problems in the world. There are three consecutive stations of the state water monitoring system. The data of each station in the data set is measured equally from upstream to downstream by the distance between stations as the id increase and the data are BOD monthly averages. The number of observations on stations is different (from 2004 to 2020). Exploratory information investigation was utilized to identify connections in the information and assess information reliance. In recent years, with the rise of artificial intelligence, more and more scholars use machine learning to solve the problem of water pollution. Both light gradient boosting machine (LightGBM) and extreme gradient boosting (XGB) use gradient boosting algorithm. XGB and LightGBM are used in this paper to build a model for predicting downstream BOD concentration and even recover some historical data that was lost. The result shows that the LightGBM determines a high accuracy model by training and testing, and LightGBM is more stable than XGB.