Prediction of Particulate Matter in a Mining Environment via Machine Learning

Abstract

The objective of this research work is to present an investigation in the application of machine learning algorithms in the prediction of particulate matter concentration in an open cast coal mine. Three types of neural networks, Long Short Term Memory, Gated Recurrent Unit and Recurrent Neural Networks are used to estimate and forecast the particulate concentration in the mine. The neural networks are bench-marked using Linear Regression and support vector machines. The results indicate that the Recurrent Neural Network performs well under conditions with short time lags, whereas both the Long Short Term Memory and Gated Recurrent Unit neural networks perform well over a wider range of lag times. Future work includes the implementation of these algorithms in early warning and intervention systems.