A Concentration Prediction and Gas Classification Model Based on LSTM-Attention Multi-task Learning Framework Network

Abstract The electronic nose (E-nose), a bionic olfactory system, has been widely used in gas identification and concentration prediction. However, these tasks are usually based on separate systems, leading to high detection costs. To address this issue, multi-task learning network model LSTM-Attention (MTL-LSTMA) as skeleton proposed simultaneously train both species recognition prediction tasks. introduction of an attention mechanism greatly reduces interference feature information, improving the pattern algorithm’s efficiency. MTL-LSTMA was tested through five sets comparison experiments, showing best performance identification. broad application prospects may revolutionize technology.