Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder that significantly impacts cognitive function and quality of life. The challenge of early and accurate diagnosis is a critical aspect in treatment of the condition, given the subtle nature of early-stage symptoms and their resemblance to other forms of dementia. This study delves onto the use of machine learning algorithms, including Support Vector Machine (SVM), Random Forest, XGBoost, and CatBoost to classify with Alzheimer’s Disease (AD), mild-cognitive impairment (MCI) and normal cognitive function (CN). The classification models utilize the extracted from radiomics features 3D MRI scans of ANDI-I dataset. The results shows that CatBoost and XGBoost outperform conventional methods. The Catboost model shows the testing accuracy of 96.3% and AUC = 0.98, indicating its ability to effectively capture intricate patterns in radiomics data. The study’s contribution to advancement of machine learning models for early diagnosis and monitoring of neurodegenerative disease is significant, as it highlights the need for further research on model interpretability and integration of multimodal data.