Python practice
Let's model airplane passengers. We'll forecast the monthly number of passengers.
This dataset is considered one of the classic time-series, published by George E.P. Box and Gwilym Jenkins alongside the book "Time-Series Analysis: Forecasting and Control" (1976). I have provided a copy of this dataset in the chapter10 folder of the book's GitHub repository. You can download it from there or use the URL directly in pd.read_csv().
We'll first start with a simple FCN and then we'll apply a recurrent network, and finally, we'll apply a very recent architecture, a Dilated Causal Convolutional Neural Network.
The FCN is first.
Fully connected network
In this first practice session, we'll use TensorFlow libraries, which we can quickly install from the terminal (or similarly from the anaconda navigator):
pip install -U tensorflow
We'll execute the commands from the Python (or IPython) terminal...
