Step by step guide to building a Deep Neural Network that classifies Images of Dogs and Cats.

Content Structure

Part 1:

1. Problem definition and Goals
2. Brief introduction to Concepts & Terminologies
3. Building a CNN Model

Part 2:

4. Training and Validation
5. Image Augmentation
6. Predicting Test images
7. Visualizing intermediate CNN layers

Problem Definition and Goals

Goal:

Build a Convolutional Neural Network that efficiently classifies images of Dogs and Cats.

Baseline Performance:

We have two classification categories — Dogs and Cats. So the probability for a random program to associate the correct category with the image is 50%. So, our baseline is 50%, which means that our model should perform well above this minimum threshold, else it is useless.

Artificial Intelligence Jobs

Dataset:

For this problem, we will use the Dogs vs Cats dataset from Kaggle, which has 25000 training images of dogs and cats combined.

#classification #ai #deep-learning #cnn