If you haven't done so yet, review the details of the dataset we are using.
Google Facets was created to explore data with machine learning in mind. It requires no coding, just loading the dataset.
It has two visualization modes (quotes from its website):
- Overview: quick understanding of the distribution of values across the features of their dataset(s). Uncover several uncommon and common issues such as unexpected feature values, missing feature values for a large number of observation, training/serving skew and train/test/validation set skew.
- Dive: an interactive interface for exploring the relationship between data points across all of the different features of a dataset. Each individual item in the visualization represents a data point. Position items by "faceting" or bucketing them in multiple dimensions by their feature values. It also allows the customization of the images it shows for each datapoint, as explained here. It is not mandatory to use custom images.
Note the disclaimer that this is not a Google product. It was developed and released as part of their work PAIR (people + AI research) with the machine learning community.
Facets Dive allows us to explore multiple dimensions of a dataset at the same time. For example, the picture below shows the distribution of the top ten diseases in the dataset by age group and gender.
To get this picture, we set the following parameters in the visualization:
- Binning | X-Axis: Finding Labels
- Binning | Y-Axis: Patient Age Group
- Color By: Patient Gender
