Table 3 All models used optimized hyperparameters.

From: Brain tumor detection with real-world predictions in Jordan hospitals

Algorithm

Hyperparameters

Neural Network

Layers: [256, 128, 64], ReLU, Dropout = 0.3, Adam(lr = 0.001), epochs = 150

SVM

kernel=’rbf’, C = 10, gamma = 0.01

Random Forest

n_estimators = 200, max_depth = 15, min_samples_leaf = 2

k-NN

n_neighbors = 7, weights=’distance’

AdaBoost

n_estimators = 200, learning_rate = 0.8

Decision Tree

max_depth = 10, min_samples_split = 5

Logistic Reg

solver=’lbfgs’, max_iter = 1000, C = 0.1