🖌 MNIST Digit Recognizer

• A simple Machine Learning Tkinter app that lets you draw digits and predicts them using a trained SVM model.

🚀 Features

• Draw digits on a canvas.

• Predict digits in real-time.

• Clear canvas anytime.

• Uses MNIST dataset and SVM model.

🛠 Technologies

• Python

• NumPy, Pandas

• Scikit-learn (SVM, Logistic Regression, Random Forest)

• Matplotlib (Confusion Matrix)

• Tkinter (GUI)

• Pillow (Image processing)

• Joblib (Save/load model & scaler)

📊 How It Works

• Train model on MNIST dataset.

• Scale data using StandardScaler.

• Save best model (SVM) and scaler.

• Tkinter GUI lets user draw digits.

• Preprocess drawing → scale → predict → show result.

⚙️ Future Ideas

• Web version with Streamlit

• Support multi-digit input

• Use CNN for higher accuracy