Dynamic monitoring and intelligent pre-alarm method for urban business environment based on spatiotemporal data fusion

Traditional business environment assessment methods mostly rely on static data and periodic surveys, so it is difficult to monitor dynamic changes and warn risks in real time. Spatio-temporal data fusion technology can integrate multi-source and heterogeneous spatio-temporal data and mine the hidden spatio-temporal characteristics and laws behind the data. In view of this, this paper presents a dynamic monitoring and intelligent pre-alarm method of urban business environment based on spatio-temporal data fusion. By building a multi-dimensional monitoring index system including economy, policy and market, and combining GIS technology and time series analysis, the temporal and spatial evolution of business environment is described. On this basis, a deep learning pre-alarm model based on LSTM is designed, and the long-term dependence is captured by means of input gate, forgetting gate and output gate mechanisms, which effectively improves the prediction accuracy. The experimental results show that the model has achieved excellent results in the business environment risk pre-alarm task, with a recall rate of over 89% and an accuracy rate of 96%, which is obviously superior to the traditional decision tree algorithm (DT).