MNIST Handwritten Digit Recognition

This video showcases a Handwritten Digit Recognition Web App built using Python, Scikit-learn, and Gradio. The model is trained on the full MNIST dataset using a Random Forest Classifier, and deployed as an interactive app where users can draw digits in real-time and receive instant predictions.

🔍 Project Highlights:
Trained on the full MNIST dataset (28x28 grayscale images)

Built and evaluated two classifiers:

SGD Classifier

Random Forest Classifier (used in the final app)

Achieved high accuracy with Random Forest on unseen test data

Visualized misclassifications and confusion matrices

Developed a Gradio-based digit drawing interface

Deployed the model locally for real-time digit prediction

✅ Tools & Libraries Used:
Python

Scikit-learn

Gradio

NumPy & Matplotlib

OpenCV

Google Colab (for training)

Joblib (for model serialization)

📁 GitHub Repository:
https://github.com/Abdulmoiz-25/MNIST-Handwritten-Digit-Recognition.git

👨‍💻 Developed by:
Abdul Moiz Meer (Intern at ARCH Technologies)
🆔 Intern-ID: 2505-0621

🎓 Internship Task:
Category B—Intermediate (Task 1)
🧠 MNIST Digit Recognition using Classification and Web App Deployment

#MNIST #DigitRecognition #MachineLearning #GradioApp #RandomForest #ClassificationModel #PythonProjects #InternshipProject #ARCHTechnologies #ScikitLearn

Видео MNIST Handwritten Digit Recognition канала Abdul Moiz Meer