Published Paper Details:

Handwritten digit recognition is an important application in computer vision, often utilized in banking, postal services, and educational tools. This paper presents a deep learning-based method that makes Convolutional neural network (CNN) use to appropriately classify the handwritten numbers. MNIST is one of the reference datasets used to train the model, which contain diverse samples of handwritten numerals. The CNN architecture includes multiple convolutional and pooling layers,Then come layers that are entirely connected and Regularization of dropouts to avoid overfitting. Techniques like data augmentation and transfer learning are applied to improve generalization and reduce computational load. Performance was assessed by comparing and implementing several deep learning models and machine learning models. With precision, recall, and F1-score values of 0.99%, the CNN model also earned the maximum accuracy of 99.25%. These outcomes show how reliable and efficient CNNs are in recognizing handwritten digits. The system exhibits enormous potential for Empirical world uses that require precise and quick digit categorization.