A Monte Carlo-Based Search Strategy for Dimensionality Reduction in Performance Tuning Parameters

Redundant and irrelevant features in dimensional data increase the complexity in the underlying mathematical models. it is necessary to conduct pre-processing steps that search for the most relevant features in order to reduce the dimensionality of the data. this work makes use of a meta-heuristic search approach that uses lightweight random simulations to balance between the exploitation of relevant features and the exploration of features that have the potential to be relevant. in doing so, this work evaluates how effective the manipulation of the search component in feature selection is on achieving a high accuracy with reduced dimensions. a control group experimental design is used in order to observe the factual evidence. the context of the experiment is the high-dimensional data experienced in the performance tuning of complex database systems. the wilcoxon signed-rank test at the .05 level of significance is used to compare the repeated classification accuracy measurements on the independent experiment and control group samples. encouraging results with a p-value < 0.05, were recorded and provided evidence to reject the null hypothesis in favour of the alternative hypothesis, which states that the meta-heuristic search approaches are effective in achieving a high accuracy with reduced dimensions depending on the outcome variable under investigation.