The notebook presents an application of deep learning that improves accuracy over logistic regression or plain-vanilla neural network. It easily achieves classification accuracy of over 80% in recognizing cat images!
What to use this notebook for:
- Access a linear algebraic framework for fast computations in Deep Learning.
- Develop intuition for how hidden layers increase a neural network's capacity for learning.
- Customize the classifier with image datasets.
Further developments in computer's ability to recognize images are found in Convolutional Neural Networks and Neural Transfer Learning.