Ensemble Prediction Algorithm of Anomaly Monitoring Based on Big Data Analysis Platform of Open-Pit Mine Slope

Table 1

Characteristics of each time series prediction’s algorithm.

Method’s names	Method’s descriptions	Advantages	Disadvantages
NARMAX	Naive prediction	Eliminate the effect of random fluctuation	Need a large number of previous data records
Autoregressive	Linear regression	Few data needed. Series of independent variables can be used for prediction	Must have autocorrelation
Neural network	Distributed parallel information processing	Functions of self-learning and associative memory	Data are easily lost. Local minimization problem
SVM	Nonlinear kernel; Linear learning algorithm	Seize key samples. Eliminate a huge number of redundant samples	Difficult to be implemented for large training samples
GEP	Heuristic algorithm	Use simple code to solve difficult problems	Easily fall into local optimum