✍️ Handwritten Character Recognition

CodeAlpha Machine Learning Internship — Task 3

📌 Objective

Identify handwritten digits and characters using a Convolutional Neural Network (CNN).

🛠️ Technologies Used

• Python 3.x
• TensorFlow / Keras
• Scikit-learn
• NumPy
• Matplotlib, Seaborn

🤖 Model Architecture

Input (28x28x1)
    ↓
Conv2D(32) → BatchNorm → Conv2D(32) → MaxPool → Dropout
    ↓
Conv2D(64) → BatchNorm → Conv2D(64) → MaxPool → Dropout
    ↓
Flatten → Dense(256) → BatchNorm → Dropout
    ↓
Dense(10) → Softmax Output

📁 Dataset

Dataset	Content	Size
MNIST	Digits 0–9	70,000 images
EMNIST (optional)	Letters A–Z	145,600 images

MNIST downloads automatically via TensorFlow.

🚀 How to Run

1. Clone this repository

git clone https://github.com/24a31a43a0/CodeAlpha\_HandwrittenCharacterRecognition

2. Install dependencies

pip install tensorflow scikit-learn numpy matplotlib seaborn

3. Open the notebook

jupyter notebook CodeAlpha\_HandwrittenCharacterRecognition.ipynb

4. Run All Cells — MNIST auto-downloads!

📈 Results

Metric	Score
Test Accuracy	~99%
Dataset	MNIST (10,000 test images)
Model Parameters	~200K

👤 Author

Venum Mery CodeAlpha ML Intern

---

Developed as part of CodeAlpha Machine Learning Internship Program