This project focuses on implementing a machine learning model for recognizing handwritten characters from A-Z. The model utilizes a neural network architecture, specifically Convolutional Neural Networks (CNN), which is trained on a dataset consisting of images representing alphabets from A-Z.
The dataset used in this project comprises 372,450 images of alphabets, each with a resolution of 28x28 pixels. The dataset plays a crucial role in training the model to recognize different characters accurately.
Convolutional Neural Networks are employed in this project to extract relevant features from the input images. The CNN model utilizes multiple layers of filters to process the images and learn distinctive patterns that represent the alphabets effectively.
Considering the large dataset size, the model is trained for a single epoch in the current implementation. However, to achieve higher accuracy in character recognition, it is recommended to train the model for multiple epochs. Training the model for longer durations allows it to learn more complex patterns and generalize better.
