Alzheimer's disease (AD) is a brain disorder that is associated with memory loss and is typically observed in elderly and aging individuals. This condition is irreversible in nature. Neural networks have shown better performance than traditional machine learning algorithms when it comes to analyzing high-dimensional data from Magnetic Resonance Images (MRI) brain images. This is an approach for processing such data. Detection of Alzheimer's disease is essential because the early treatment can control the progression of the disease in a slower pace. In this work, the proposed model used different models of convolutional neural networks (CNN) techniques to process the MRI images of the brain and classified them into different classes. The proposed model detects AD using neural network feature extraction and classification methods based on the pre-processed datasets. Here, in the proposed model, various algorithms such as Resnet-50, DenseNet-169 and Hyperparameter tuning of Convolutional Neural Network are utilized to find the better performing model in terms of accuracy when compared with other traditional techniques.