Digit-and-Character-Recognition-with-LIME

This project focuses on developing machine learning models for recognizing handwritten digits and characters using the MNIST and EMNIST datasets. It also explores model interpretability through the use of LIME (Local Interpretable Model-agnostic Explanations).

Digit Recognition with LIME (97% Accuracy) and Character Recognition

Features

• Handwritten Digit Recognition: Models trained on the MNIST dataset for accurate digit classification.
• Character Recognition: Leveraged the EMNIST dataset for recognizing a variety of characters.
• Model Interpretability: Utilized LIME to provide visual explanations, offering insights into the decision-making process of the models.

Datasets

• MNIST Dataset: Contains 60,000 training images and 10,000 testing images of handwritten digits (0-9).
• EMNIST Dataset: Provides a variety of handwritten characters, including letters and digits, across multiple splits.