Chapter 3. Introduction to neural prediction: forward propagation

“I try not to get involved in the business of prediction. It’s a uick way to look like an idiot.”

Warren Ellis comic-book writer, novelist, and screenwriter

Step 1: Predict

This chapter is about prediction

In the previous chapter, you learned about the paradigm predict, compare, learn. In this chapter, we’ll dive deep into the first step: predict. You may remember that the predict step looks a lot like this:

In this chapter, you’ll learn more about what these three different parts of a neural network prediction look like under the hood. Let’s start with the first one: the data. In your first neural network, you’re going to predict one datapoint at a time, like so:

Later, you’ll find that the number of datapoints you process at a time has a significant impact on what a network looks like. You might be wondering, “How do I choose how many datapoints to propagate at a time?” The answer is based on whether you think the neural network can be accurate with the data you give it.

A simple neural network making a prediction

What is a neural network?

What does this neural network do?

Making a prediction with multiple inputs

Multiple inputs: What does this neural network do?

Multiple inputs: Complete runnable code

Making a prediction with multiple outputs

Predicting with multiple inputs and outputs

Multiple inputs and outputs: How does it work?

Predicting on predictions

A quick primer on NumPy

Summary