The data that we've looked at so far is known as static data. It doesn't contain information that can be varied through the time frame dynamically. However, it is also necessary for us to deal with the data changing. Examples of this include audio data and natural language. Their major characteristic is the fact that each point depends on the previous points in the sequence. While there are supervised learning techniques that predict labels by considering the dependencies within the sequence, we are going to focus on the underlying structure of the sequence.
In this chapter, we are going to take a look at techniques we can use to analyze sequential data. Specifically, we will cover Fourier transformation and its implementation in TensorFlow.js.
The following topics will be covered in this chapter:
- What is Fourier transformation?
- Cosine curve...
