| | Failure rate prediction method | Single model | Statistical model | Regression analysis [1], time series [2, 3], mathematical statistics [4], Weibull distribution statistics [5], Bayesian [6] | | Grey model | GM (1, 1) [7–9], Verhulst [10] | | Machine learning model | Artificial neural network (ANN) [11], BP neural network [12–14], generalized regression neural network (GRNN) [15], support vector machine (SVM) [16], least squares support vector machine (LS-SVM) [17], random forest [18] | | Deep learning model | Long short-term memory (LSTM) [19], convolutional neural network (CNN) [20] | | Combined model | Model-based combination forecasting | Grey neural network-fuzzy recognition [21], artificial neural network and genetics [22], MLR-GM (1, N)-PLS-BP-SVM [23], SVR- multiple regression-principal component analysis [24], ARMA-BP [25], grey model combination [26, 27] | | Method-based combination model | Holt-winters seasonal model [28], neural network residual correction AR [29], artificial neural network Weibull regression [30], Weibull-based generalized renewal process (WGRP) [31], sparse direct support vector machine regression [32], generalized weighting least-squares combination [33] | | Integrated combination model based on decomposition | Empirical mode decomposition (EMD) and LS-SVM combination [34], correlation vector EMD and GMDH combination [35], EMD and RVM-GM combination [36], CEEMD and combined model [37] |
|
|
