Interpreting your AutoML results

Your training run should have taken about 15 minutes and produced a model with around 80% accuracy. However, there is much more to your results than this simple metric. There are data guardrails that will inform you of any problems with your data. There is also a slew of different metrics for each of the three problem types and accompanying charts and graphs that can assist you in presenting your results to the business:

1. To begin, click Automated ML from the left-hand menu of AMLS and click the latest run from your Titanic-Training experiment, as seen in Figure 3.12:

   Figure 3.12 – Examining your results

   You will be taken to a screen with a variety of metrics regarding your model, including the type of algorithm used to train the best-performing model, its accuracy score, the date and time it was created, and how long your AutoML run took to execute. Take advantage of the Description area in the bottom-right corner of your screen to write...