COMPANY : CODTECH IT SOLUTIONS

NAME : NAVEEN K

INTERN ID : CT06DA721

DOMAIN : MACHINE LEARNING

DURATION : 6 WEEKS

MENTOR : NEELA SANTHOSH KUMAR

Task 3 – CNN for Digit Classification (MNIST)

Internship Project - Machine Learning (CodTech)

Hello! This repository contains the solution to Task 3 of the CodTech Machine Learning Internship. The goal of this task was to build a Convolutional Neural Network (CNN) using TensorFlow/Keras to classify handwritten digits (0–9) from the MNIST dataset.

The deliverable includes:

• A functional deep learning model
• Performance evaluation on a test dataset
• A saved .h5 model file
• Instructions to use the model for prediction on new images

---

Objective

To build a CNN that:

• Takes in 28x28 grayscale images of handwritten digits
• Learns to classify them into one of 10 classes (0–9)
• Evaluates performance on a test dataset
• Can be saved and reused for new predictions

---

Dataset: MNIST

The MNIST dataset is a classic benchmark in machine learning:

• 60,000 training images
• 10,000 test images
• Grayscale format, size: 28x28 pixels
• Labels: Digits from 0 to 9

Sample MNIST:

Accuracy and Loss Epochs:

Confusion Matrix: