A New Coal Seam Gas Content Prediction Model Based on Feature Selection and Machine Learning Fusion in Sparse Datasets

Abstract

Coal seam gas content is a key factor for mine safety and methane resource utilization. To improve prediction accuracy, we collected gas content data from 27 groups on the left flank of the Huaibei Taoyuan Coal Mining area. The main factors affecting gas content were selected based on geological structure, coal quality characteristics, and the conditions of coal occurrence. Subsequently, various algorithms were used to identify the main controlling factors of gas content, and different regression prediction models were applied based on the selected feature combinations. The Genetic Algorithm with Ant Colony Optimization (GA-ACO) was introduced and compared with several traditional feature selection algorithms, including All Subset Regression (ASR), Forward Selection Regression (FSR), Backward Selection Regression (BSR), Random Forest Algorithm (RFA), and Lasso regression, to construct the main control factor set for gas content. Additionally, The Particle Swarm Optimization Support Vector Regression (PSO-SVR) to build a regression prediction model. Various regression models such as Multivariate Linear Regression (MLR), Multi-Layer Perceptron (MLP), Gradient Boosting Regression Tree (GBRT) and SVR were compared to verify the reliability of the model. The dataset was randomly divided into training and validation sets in an 8:2 ratio, and the prediction performance was evaluated using metrices. The results show that the use of feature selection and machine learning effectively predicts the gas content of coal seams. PSO-SVR model based on ASR achieved the best prediction performance with an R2of 0.99 on the training set and 0.978 on the test set. The maximum error was 0.077. Finally, Using the constructed model to predict regional gas content and plotted a regional gas content distribution map. This study demonstrates the potential of combining feature selection with machine learning for regression prediction in coal bed geology, offering a feasible and promising approach for research in this field.

Keywords: Coal and gas outburst, Coal seam gas content, GA-ACO algorithm, PSO-SVR model