Chapter 3: Training Models with Tabular Data

In the previous chapter, we learned how to ingest various kinds of datasets using fastai and how to clean up datasets. In this chapter, we are going to get into the details of training a model with fastai using tabular data. Tabular data, which is data organized in rows and columns that you would find in a spreadsheet file or a database table, is critical to most businesses. The fastai framework acknowledges the importance of tabular data by providing a full suite of features to support deep learning applications based on tabular data.

To explore deep learning with tabular data in fastai, we will return to the ADULT_SAMPLE dataset, one of the datasets we examined in Chapter 2, Exploring and Cleaning Up Data with fastai. By using this dataset, we will train a deep learning model, while also learning about the TabularDataLoaders (used to define the training and test datasets) and tabular_learner (used to define and train the model) objects...