Abstract: The study of gas outflow from coal tunnel working face is of great significance for the prevention and control of gas in coal tunnel working face. Using the characteristics of deep learning theory and long and short-term memory neural network to process the time series samples efficiently, a prediction model of gas outflow prediction model based on LSTM neural network is established. The hyperparameters of the model are optimised according to the size of the loss value in the training process, and the optimal hyperparameters of the model are selected and determined. With the help of the original data of gas outflow from the coal mine face, the applicability and accuracy of the model are verified, and the variation trend of gas emission in time dimension is analyzed according to the predicted results. The results of the study are of reference significance for predicting the trend of gas outflow in coal mine face, identifying abnormal gas outflow in the face, and improving the level of gas control in the face.