View source: R/model_building.R
Function basically performing the same task as e1071::tune, but here made clear.
1 2 3 4 | select_svm(input_frame, kernel = "linear", max.polynomial = 2,
cachesize = 500, min_lag = 7, eps_grid = 10^c(0.5, 0.75, 1, 1.25, 1.5),
gamma_grid = 10^c(0.5, 0.75, 1, 1.25, 1.5), sigma_grid = -10^seq(-1, 1,
0.5), cores = 8L)
|
input_frame |
A data frame containing a date and a price. |
cachesize |
cache memory in MB (default 4000) |
max.cost |
An integer containing the maximum allowed cost. If left as NULL, it will default to floor(max(abs(price)) + 3 * sd(price)) |
max.epsilon |
An interger being the max allowed epsilon parameter. If left as null, it will default to 4, being the value where no support vectors are returned for the base spot price for Germany. |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.