There are several ways that you can learn new ideas and learn new skills. In an art class students study colors, but aren't allowed to actually paint until college. Sound absurd?
Unfortunately, this is how most modern machine learning is taught. The experts are doing something similar. They tell you that need to know linear algebra, calculus and deep learning. This is before they'll teach you how to use natural language Processing (NLP).
In this book, I want us to learn by teaching the the whole game. In every section, we see how to solve real-world problems and learn the tools along the way. Then, we will dig deeper and deeper into understanding how to make these toolks. This learning and teaching style is very much inspired by Jeremy Howard of fast.ai fame.
The next focus is to have code examples wherever possible. This is to ensure that there is a clear and motivating purpose behind learning a topic. This helps us understand with intuition, beyond math formulae with algebraic notation.
In this opening chapter, we will focus on an introduction to NLP. And, then jump into a text classification example with code.
This is what our journey will briefly look like:
- What is NLP?
- What does a good NLP workflow look like? This is to improve your success rate when working on any NLP project.
- Text classification as a motivating example for a good NLP pipeline/workflow.
