Deformation prediction of landslide based on functional network

This paper proposes functional networks as novel intelligence paradigm scheme for landslide displacement prediction. They evaluate unknown neuron functions from given functional families during the training process. General functional networks with two variables training data set (GFN), separable functional networks (SFN) and associativity functional networks (AFN) are applied to forecast a real-world example. In addition, we compare them with back-propagation neural network (BPNN) in terms of the same measurements. The results reveal that the landslide displacement prediction using functional networks is reasonable and effective, and GFN are consistently better than the other two types of functional networks and BPNN.