We utilize Generative Adversarial Networks (GANs) to generate real and synthetic cat images. Subsequently, we apply Principal Component Analysis (PCA) and AutoEncoders for dimensionality reduction. Finally, we employ Gradient Boosting to assess the model's accuracy