A Multi-Dimensional Time Series Data Mining Model for Weather Forecast
Kun Liu · Advanced materials research · 2012
Based on the analysis and actual application under the background of the ground automatic weather stations research material, through selecting dimension, piecewise linear fitting method, the cluster symbol, this thesis proposes the dimensions redundant reduction algorithm, the extremum slope piecewise linear fitting method, a multi-dimensional time series data mining model for the meteorological data, and uses the model to preliminarily mine the rule of rain weather phenomena. Finally, the experimental results show that this model designed in this paper can predict the weather phenomenon great practicality.