Gliomas Disease Prediction: An Optimized Ensemble Machine Learning-Based Approach

The most frequent primary brain tumors are gliomas, which call for precise prognostic models for early detection and individualized care. For optimum treatment planning and individualized patient care, accurate prediction of glioma development and patient survival is essential. For improving treatment choices and patient outcomes in glioma care, the convergence of these techniques offers enormous promise. The multifaceted field of glioma prediction integrates many modalities and machine-learning strategies to increase prognostication accuracy. To assist in the precise and efficient prediction of glioma tumor development, grade, and patient prognosis, this research study provides a glioma prediction model that makes use of machine learning algorithms that are KStar and SMOreg. In this research, a voting-based approach is introduced aimed at enhancing the performance outcomes of both the feature selection phase and machine learning models employed in the prediction of glioma. This study incorporates machine learning methods for glioma prediction. To determine the optimal scheme for the selected ensemble learning model in a voting-based approach, various techniques are employed to identify the most effective option. The publicly available TCGA dataset with 24 attributes and 839 instances. The computational results indicate that the proposed method achieves 96.3% accuracy on the TCGA dataset. The suggested glioma prediction model exhibits encouraging findings and has great promise for guiding clinical judgments and enhancing patient outcomes.