RNN for Time Series Data with TensorFlow and Keras

Time series data is a sequence of values, recorded or measured at different time intervals. Being a sequence, the RNN architecture is the best method to train models from such data. In this chapter, we will use a sample time series data set to showcase how to use TensorFlow and Keras to build RNN models.

We will cover the following topics in this chapter:

• Airline passengers time series dataset:
  • Description and downloading of the dataset
  • Visualizing the dataset
• Preprocessing the dataset for RNN in TensorFlow
• RNN in TensorFlow for time series data:
  • SimpleRNN in TensorFlow
  • LSTM in TensorFlow
  • GRU in TensorFlow
• Preprocessing the dataset for RNN in Keras
• RNN in Keras for time series data:
  • SimpleRNN in Keras
  • LSTM in Keras
  • GRU in Keras

Let's start by learning about the sample dataset.

You can follow along with the code...