Weekly Challenges

AYXAcademy · ‎07-17-2023

A solution to last week's challenge can be found here.

This week's challenge marks the culmination of a trilogy of challenges inspired by the 2023 Inspire Grand Prix. These challenges explore real-world scenarios that many companies frequently face. The previous week's challenge involved a spatial problem, while the week prior focused on data preparation and integration. In this concluding challenge, we will delve into a predictive case.

If you are eager to experience the same exhilaration our racers feel in Las Vegas, take a quick, 2-minute glance at the instructions, start your timer, and record how long it takes you to determine the correct answer! Remember to share your time when you submit your workflow.

Let’s start now: 3, 2, 1, Go!

A driver who works for ACE was recently promoted to shift manager, so the company needs to reduce their weekly food collection schedule by five shifts until they can fill the driver position.

1. Which five shifts are most likely to be unsuccessful?

Build a random forest model to determine the five collection shifts with the highest likelihood of having a cancellation based on historical job data. For the forest model, use only the DistanceMiles, Hour, and DayofWeek columns as your predictors.

For the cancellation data, you will need the calculate the following:

DayofWeek: The exact field name, full name of the day, and (%A) the job occurred based on the Date.

JobStatus:

• A value of Unsuccessful if the job was canceled.
• A value of Successful if the job was successful based on ClosedReason values. A Successful value would be any ClosedReason starting with 01 or 02 and an Unsuccessful value would be any other value.