Training Data for Machine Learning

Chapter 6. Theories, Concepts, and Maintenance

Introduction

So far, I have covered the practical basics of training data: how to get up and running and how to start scaling your work. Now that you have a handle on the basics, let’s talk about some more advanced concepts, speculative theories, and maintenance actions.

In this chapter I cover:

• Theories

• Concepts

• Sample creation

• Maintenance actions

Training a machine to understand and intelligently interpret the world may feel like a monumental task. But there’s good news; the algorithms behind the scenes do a lot of the heavy lifting. Our primary concern with training data can be summed up as “alignment,” or defining what’s good, what should be ignored, and what’s bad. Of course, real training data requires a lot more than a head nod or head shake. We must find a way to transform our rather ambiguous human terminologies into something the machine can understand.

A note for the technical reader: This chapter is also meant to help form conceptual understandings of the relationships of training data to data science. The data science technical specifics of some of the concepts brought up here are out of the scope of this book, and the mention of the topics is only in relation to training data, not an exhaustive account.

Theories

There are a few theories that I think will help you think about training data better.

I’ll introduce the theories here as bullet points, and then each one will be explained in each section:

• A system is ...