A beautiful collection of some past FullRank ML applications.
Utilities for quickly visualizations in 3D are available.
Over the course of several iterations, the loss (in this case average percent) in reduced.
Using the AutoML tooling, the features are ranked in order of their contribution to the average loss reduction. This process is iteratively done until.
Compute and display feature correlations.
This chart depicts the close ridings in a recent Canadian election.
A list of items can be visually displayed using a variety of tree structures.
A clustering of 400+ home sales in the Greater Toronto Area (GTA) by location.
FullRank has a 'FullReports' feature that can automatically create detailed pdf reports for distribution.
A file browser has been implemented for picking files.
Browse different version of hypotheses.
This tool allows one to dynamically create hypotheses and test the results interactively.
In this Hypothesis Prospector panel, we see a summary of the predicted value for one observation.
This plot quickly displays the 'information' available in many columns at a glance.