Select landslide susceptibility main affecting factors by multi-objective optimization algorithm

Landslide often causes great damage to human society. Landslide susceptibility and hazard zoning is an efficient way to reduce landslides risk. There are many factors make contribution to landslide susceptibility. Artificial neural network has been devoted to this area but still need to improve its effectiveness. This paper present a strategy mixed with evolutionary algorithm and neural network. After training on historical data with neural network, the objective function could be replaced by feasible neural network mode. Then, evolutionary computation could be followed by embedding the given neural network model in multi-objective evolutionary algorithm. The given method was applied to Miyi county, southwest China. Its result have take into practice and proved effective.