Standard Genetic Algorithm. Implements a standard genetic algorithm using GA package (ga) with a fitness function specialised for feature selection.

geneticalgorithm(
  model = feamiR::svmlinear,
  k = 30,
  training,
  test,
  parallel = T,
  mutprob = 0.1,
  crossprob = 0.8,
  popsize = 20,
  maxiter = 1000,
  maxiter_withoutimprovement = 300,
  numberpassedon = 3,
  plot = F
)

Arguments

model	The ML models used to classify the data, typically SVM with a given kernel
k	Maximum number of features to be output.
training	Training dataset as a data.frame with classification column and column for each feature.
test	Test dataset with matching columns to training.
parallel	Specifies whether GA should be run sequentially or in parallel (default: T)
mutprob	The probability that an individual undergoes mutation in a particular iteration (default: 0.1)
crossprob	The probability of crossover between pairs of individuals (default: 0.8)
popsize	The size of the solution population (default:20)
maxiter	The maximum number of iterations to run before termination (default: 1000)
maxiter_withoutimprovement	The maximum number of consecutive iterations without improvement to fitness before termination (default: 300)
numberpassedon	The number of best fitness individuals to be passed on to the next generation in each iteration (default: 3)
plot	Specifies whether GA plot should be shown (default: F)

Value

Set (unordered) of <=k features and training and test accuracy, sensitivity and specificity using these features.

Examples

data_train = data.frame(
      classification=as.factor(c(1,1,0,0,1,1,0,0,1,1)),
      A=c(1,1,1,0,0,0,1,1,1,0),
      B=c(0,1,1,0,1,1,0,1,1,0),
      C=c(0,0,1,0,0,1,0,0,1,0),
      D=c(0,1,1,0,0,0,1,0,0,0),
      E=c(1,0,1,0,0,1,0,1,1,0))
data_test = data.frame(
      classification=as.factor(c(1,1,0,0,1,1,1,0)),
      A=c(0,0,0,1,0,0,0,1),
      B=c(1,1,1,0,0,1,1,1),
      C=c(0,0,1,1,0,0,1,1),
      D=c(0,0,1,1,0,1,0,1),
      E=c(0,0,1,0,1,0,1,1))
geneticalgorithm(
  feamiR::svmlinear,
  k=2,
  data_train,
  data_test,
  parallel=FALSE,
  maxiter=5,
  maxiter_withoutimprovement=5,
  popsize=10)
#> $feature_list
#> [1] "C" "D"
#> 
#> $fitness
#> [1] 0.7
#> 
#> $testaccuracy
#> [1] 0.875
#> 
#> $trainspec
#> [1] 0.5
#> 
#> $testspec
#> [1] 0
#> 
#> $trainsens
#> [1] 0.8333333
#> 
#> $testsens
#> [1] 0.8
#>