Prediction of sea surface temperature in the tropical Atlantic by support vector machines

An Ecological Model for Quantitative Risk Assessment for Schistosomiasis: The Case of a Patchy Environment in the Coastal Tropical Area of Northeastern Brazil

“…The scenarios hold rainfall and sanitation at their initial levels for the entire 20-year projection. An improved rainfall model is being developed using support vector machines (such as those made for the prediction of sea surface temperature (36) ) that considers annual shifts in rainfall. This improvement will make our model output for PG more realistic.…”

Computing confidence and prediction intervals of industrial equipment degradation by bootstrapped support vector regression

“…The substitution of the dot products in Eq. (8) by (13) does not affect the way of solving the dual problem, hence the dot products in the estimated regression expression can be replaced by the RBF kernel as well.…”

“…These methods do not require specific knowledge on the functional relationship between the influential factors (the explanatory variables) and the degradation variable of interest. Among these methods, Support Vector Regression (SVR) has provided promising results in reliability [8,9], economic [10,11], Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/ress environmental [12,13], electrical [14,15] applications, among others. The SVR learning (or training) phase involves the resolution of a convex quadratic optimization problem, for which the Karush-Kuhn-Tucker (KKT) first order conditions are necessary and sufficient for a global optimum [16].…”

Prediction of the binary surface tension of mixtures containing ionic liquids using Support Vector Machine algorithms

“…is the kernel function that efficiently handles the non-linear relations between x and y [96]. By solving the optimization problem of Eq.…”

“…where C is a user pre-defined positive real constant which measures the trade-off between model's capacity to predict unseen data and training accuracy, w is the vector of coefficients, b is the bias term, ξ i ; ξ Ã i are flexible variables governing the training errors, and ϵ, a positive adjustable parameter, is the tube width of the Vapnik's ε-insensitive loss function controlling the regression precision [52,96,97], as shown in Fig. 2.…”