Challenge #127: US Grand Prix Lap 2 - Employee Retention

[kelly_gilbert]

I'm in the People Analytics field, so this was up my alley!

Spoiler

I do want to point out that we would usually train the model using a subset of the data (e.g. the state of the workforce at the end of a prior year), rather than both training and scoring the model using the same set of current employees.

[jssandom]

Model output seems to be different in this version but here's my solution!

[pasccout]

Here is my solution... took me 15-20 minutes to figure it out...

the problem is that the results are not good... I did download solutions (After) from others that look similar and they also don't provide the correct join... any special config I need to do?

[jamielaird]

Here's my solution. I was only able to replicate the desired output after I changed the Decision Tree back to version 1.0.

Spoiler

[RolandSchubert]

My solution - I had to change the tool version for the Decision Tree tool to 1.0 as well,
using the most current version returns no records with X_Terminated >= 0.5

[Natasha]

Spoiler

[ipeng]

Like to do prediction

[JoshuaGostick]

My solution :)

Spoiler

[cplewis90]

Spoiler

[KMiller]

The solution is attached. This is weird though, my solution is the same as everyone else's but no employees have >0.5 leaving probability. I have downloaded a few other people's solutions and when I run them they do not have >0.5 probabilities either. I wonder if the latest version has changed the basis of calculation for the decision tree tool.