Explainable time–frequency convolutional neural network for microseismic waveform classification

•The XTF-CNN provides both excellent learning ability and explain-ability for time series classification.•The dual-channel neural network TF-CNN learns high-level features of time series in the both time and frequency domains.•The explanation strategy EUG-CAM provides comprehensible explanation to the classification results.•The performance is evaluated by using real-world microseismic wave-forms collected from an underground project.