sits_gbm: Train SITS classifiction models with Gradient Boosting...

In luizassis/sits: Satellite Image Time Series Analysis

Description

Use generalized liner model (glm) via penalized maximim likelihood to classify data. This function is a front-end to the "cv.glmnet" method in the "glmnet" package. Please refer to the documentation in that package for more details.

Usage

sits_gbm(distances.tb = NULL, formula = sits_formula_logref(),
  distribution = "multinomial", n.trees = 5000, interaction.depth = 4,
  shrinkage = 0.001, cv.folds = 5, n.cores = 1, ...)

Arguments

distances.tb	a time series with a set of distance measures for each training sample
formula	a symbolic description of the model to be fit. SITS offers a set of such formulas (default: sits_formula_logref)
distribution	the name of the distribution. Either "multinomial" for classification)
n.trees	Number of trees to fit. This should not be set to too small a number, to ensure that every input row gets predicted at least a few times. (default: 500)
interaction.depth	The maximum depth of variable interactions. 1 implies an additive model, 2 implies a model with up to 2-way interactions.
shrinkage	a shrinkage parameter applied to each tree in the expansion. Also known as the learning rate or step-size reduction.
cv.folds	number of cross-validations to run
n.cores	number of cores to run
...	other parameters to be passed to 'gbm::gbm' function

Value

result either an model function to be passed in sits_predict or an function prepared that can be called further to compute multinom training model

Author(s)

Alexandre Xavier Ywata de Carvalho, [e-mail]

Rolf Simoes, [e-mail]

luizassis/sits documentation built on May 30, 2019, 7:15 p.m.