Comparative Evaluation Of Machine Learning Classifiers For Brain Tumor Detection

Umair Ali
DOI: https://doi.org/10.1101/2024.07.28.24311114
2024-08-12
Abstract:This study evaluates the effectiveness of six machine learning classifiers - Support Vector Classifier (SVC), Logistic Regression, K-Nearest Neighbors (KNN), Naive Bayes, Decision Tree, and Random Forest - in detecting brain tumors using numerical data rather than traditional imaging techniques like MRI. The results emphasize the importance of data preprocessing, particularly feature scaling, in enhancing model performance. Among the classifiers, Random Forest emerged as the top performer, achieving an accuracy of 98.27% on both original and scaled data, demonstrating its robustness and reliability. The study highlights the potential of Random Forest as a valuable tool for automated brain tumor detection in clinical settings, offering a cost-effective and accessible alternative for resource-constrained environments. The paper suggests that future research should explore advanced deep learning models, such as 3D Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs), to further improve diagnostic accuracy and support early intervention and personalized treatment strategies for brain tumor patients.
Oncology
What problem does this paper attempt to address?
The main objective of this paper is to evaluate the effectiveness of several machine learning classifiers in detecting brain tumors using numerical data instead of traditional Magnetic Resonance Imaging (MRI) images. Specifically, the study examines the performance of the following six classifiers: 1. **Support Vector Classifier** (SVC) 2. **Logistic Regression** 3. **K-Nearest Neighbors Algorithm** (KNN) 4. **Naive Bayes Classifier** 5. **Decision Tree Classifier** 6. **Random Forest Classifier** The paper emphasizes the importance of data preprocessing, particularly the role of feature scaling in enhancing model performance. The study results show that the Random Forest Classifier performed the best on both raw and scaled data, achieving an accuracy of 98.27%, demonstrating its robustness and reliability. The study also proposes a cost-effective and easily accessible method to assist in the automatic detection of brain tumors in clinical settings, particularly suitable for resource-limited environments. Additionally, the paper suggests that future research could explore more advanced deep learning models, such as 3D Convolutional Neural Networks (3D CNNs) and Generative Adversarial Networks (GANs), to further improve diagnostic accuracy and support early intervention and personalized treatment strategies. In summary, the problem this paper attempts to address is how to effectively detect brain tumors using numerical data and identify the most effective machine learning classifier for this task. The study's findings not only help improve brain tumor diagnostic techniques but also provide valuable directions for subsequent research.