Optimizing Brain Tumour Classification: A Fusion Approach of PCA, SVM, and Image-Based Features

This article highlights a novel method for improving the precision and efficacy of brain tumour classification. The proposed method combines image-based features with Principal Component Analysis (PCA) and Support Vector Machine (SVM) classifiers. The application of preprocessing techniques ensures the consistency and quality of the data. PCA is used to reduce dimensionality while preserving informational distinctions. The RBF kernel is utilised by SVM classifiers trained separately on PCA-transformed and original image features. The classification of brain tumours is more precise when measured using standard metrics than when compared to conventional methods. A fusion strategy combines decision functions by aggregating the probability scores of each classifier, resulting in predictions that are robust and accurate. The results of the proposed model makes it easier and more accurate to diagnose brain tumours. Clinical applications of this integration of PCA and SVM classifiers with image-based features hold promise for medical image analysis.classification over existing state-of-the-art techniques.