The method proposed here under the name ‘Heuristic search’ is currently under review for a journal with the title ‘An effective heuristic for developing hybrid feature selection in high dimensional and low sample size datasets’. In this study, an effective feature selection method that combines gradual permutation filtering with a heuristic tribrid search strategy specifically tailored for high-dimensional and low sample sizes (HDLSS) contexts is introduced. The proposed approach considers inter-feature interactions and leverages feature rankings during the search process. For more details, please refer to the paper.
Please note that the heuristic search utilizes the following package. Ensure it is installed before proceeding.
library(parallel)
library(ranger)
library(xgboost)
library(caret)
library(pROC)Using the downloaded library, perform feature selection as shown in the following example:
# Ensure the library path is set correctly. ex) 'path/heuristic_search.R'
# This example assumes it exists within the working directory.
source('heuristic_search.R')
# Load Dataset
# Assume the 'alon data' exists in the working directory.
data_path <- 'alon.csv'
original_ds <- read.csv(data_path)
# Separate features and labels
# Assume the first column of the dataset contains label information.
ds <- original_ds[,-1]
cl <- original_ds[,1]
# Perform feature selection
result <- heuristic_feature_selection(ds, cl)
print(result)