One common noninvasive diagnostic method for imaging the brain is Magnetic Resonance Imaging (MRI). Many lives might be saved by accurately analyzing brain MRI scans which aid in the early discovery of Alzheimer’s. However, a clinical and technological perspective accurately determines the state of brain MRI which includes a considerable amount of redundant information. The data’s dimensionality would rise because of redundant information. Consequently, the time and computational intricacy of the classifiers for differentiating the brain magnetic resonance images would be decreased by employing a feature selection technique to identify the ideal collection of features. This investigation aims to inspect how well feature selection performs using meta-heuristic search algorithms. The statistical characteristics were taken from the brain MRI images are subjected to feature selection utilizing Principal Component Analysis and Discrete Wavelet Transform. High-performance feature selection is demonstrated by our results.