Understanding the world in four applications: tables, text, recommender systems, and images
In their seminal paper describing fastai, Howard and Gugger (https://arxiv.org/pdf/2002.04688.pdf) describe the four application areas that fastai supports out of the box. In this section, we will go through these four applications of deep learning that fastai directly supports: tabular data, text data, recommender systems, and computer vision. The MNIST example that you saw in the previous section is an example of a computer vision application. The MNIST example included the following:
- Curated dataset: MNIST. You can find an overall list of curated datasets here:
- Easy ingestion of the curated dataset via
untar_data() - Image-specific handling of the dataset via a data loader object
- Definition of an image-specific model structure via a
Learnerobject - Utilities to examine the dataset
Similarly, fastai also provides components specifically...
