Abstract: The application of Fuzzy mathematics plays an important role in seismology, especially in earthquake prediction. Presently, a new fuzzy system method in fuzzy mathematics-the Fuzzy system by learning from examples (FSLE) is developed, which has been widely used in artificial intelligence, machine learning, series modelling and synthetical prediction, and obtained significant effect. Based on this, the authors introduced this method into trend prediction and pre-warning study of seismic activity: Firstly, the theory of fuzzy system by learning from examples (FSLE) is introduced, and then this method is firstly used in seismic prediction and pre-warning of time-series of the earthquake maximum magnitude in northern China. The result shows that the FSLE method has good prospect for forecast and pre-warning of earthquakes, and so it has shown that the method may be a new effective method in analyzing of forecast and pre-warning of seismic activity.

Key words: Fuzzy system by learning from examples (FSLE), Seismic activity, Earthquake forecast, Seismic pre-warning