Practical Deep Learning for Cloud, Mobile, and Edge

A Crash Course in Convolutional Neural Networks

In keeping with the word “practical” in the book’s title, we’ve focused heavily on the real-world aspects of deep learning. The goal of this appendix is meant to serve as reference material, rather than a full-fledged exploration into the theoretical aspects of deep learning. To develop a deeper understanding of some of these topics, we recommend perusing the “Further Exploration” for references to other source material.

Machine Learning

• Machine learning helps learn patterns from data to make predictions on unseen data.

• There are three kinds of machine learning: supervised learning (learning from labeled data), unsupervised learning (learning from unlabeled data), and reinforcement learning (learning by action and feedback from an environment).

• Supervised learning tasks include classification (output is one of many categories/classes) and regression (output is a numeric value).

• There are various supervised machine learning techniques including naive Bayes, SVM, decision trees, k-nearest neighbors, neural networks, and others.

Perceptron

• A perceptron, as shown in Figure -1, is the simplest form of a neural network, a single-layered neural network with one neuron.

• A perceptron calculates a weighted sum of its inputs; that is, it accepts input values, multiplies each with a corresponding weight, adds a bias term, and generates a numeric output.

  Figure -1. An example of a perceptron

• Because a perceptron is governed ...