scorecardModelUtils (version 0.0.1.0)

support_vector_parameters: Hyperparameter optimisation or parameter tuning for Suppert Vector Machine by grid search

Description

The function runs a grid search with k-fold cross validation to arrive at best parameter decided by some performance measure. The parameters that can be tuned using this function for support vector machine algorithm are - kernel (linear / polynomial / radial / sigmoid), degree of polynomial, gamma and cost. The objective function to be minimised is the error (mean absolute error / mean squared error / root mean squared error). For the grid search, the possible values of each tuning parameter needs to be passed as an array into the function.

Usage

support_vector_parameters(base, target, scale = T, kernel, degree = 2,
  gamma, cost, error = "rmse", cv = 1)

Arguments

base

input dataframe

target

column / field name for the target variable to be passed as string (must be 0/1 type)

scale

(optional) logical vector indicating the variables to be scaled (default value is TRUE)

kernel

an array of kernels to be iterated on; kernel used in training and predicting, to be cheosen among "linear", "polynomial", "radial" and "sigmoid"

degree

(optional) an array of degree of polynomial to be iterated on; parameter needed for kernel of type "polynomial" (default value is 2)

gamma

an array of gamma values to be iterated on; parameter needed for all kernels except linear

cost

an array of cost to be iterated on; cost of constraints violation

error

(optional) error measure as objective function to be minimised, to be chosen among "mae", "mse" and "rmse" (default value is "rmse")

cv

(optional) k vakue for k-fold cross validation to be performed (default value is 1 ie. without cross validation)

Value

An object of class "support_vector_parameters" is a list containing the following components:

error_tab_detailed

error summary for each cross validation sample of the parameter combinations iterated during grid search as a dataframe

error_tab_summary

error summary for each combination of parameters as a dataframe

best_kernel

kernel parameter of the optimal solution

best_degree

degree parameter of the optimal solution

best_gamma

gamma parameter of the optimal solution

best_cost

cost parameter of the optimal solution

runtime

runtime of the entire process

Examples

# NOT RUN {
data <- iris
suppressWarnings(RNGversion('3.5.0'))
set.seed(11)
data$Y <- sample(0:1,size=nrow(data),replace=TRUE)
svm_params_list <- support_vector_parameters(base = data,target = "Y",gamma = 0.1,
                   cost = 0.1,kernel = "radial")
svm_params_list$error_tab_detailed
svm_params_list$error_tab_summary
svm_params_list$best_kernel
svm_params_list$best_degree
svm_params_list$best_gamma
svm_params_list$best_cost
svm_params_list$runtime
# }