Single-step and Multi-step Time Series Prediction for Urban Temperature Based on LSTM Model of TensorFlow
Wen Yue Zhang, Jia Feng Xie, Guo Chun Wan, Mei Song Tong · 2021
In this paper, a temperature prediction model based on long - term memory network (LSTM) is developed by modeling urban temperature data. The fluctuations of temperature changes are usually nonstationary fluctuations and many meteorological factors need to be considered in the analysis. At the same time, the temperature data is very sensitive to time series. The LSTM network is often used to deal with time series data problems, and has the advantage of being able to remember information for a long time to predict temperature. The model based on TensorFlow is mainly introduced to predict the future temperature by learning the historical temperature samples. Compared with other time series data prediction methods, it has more accurate prediction effect.