Challenge #44: Inspire Europe '16 Grand Prix (L3)

[GeneR]

The link to last week’s challenge (challenge #43) is HERE

As a follow on from last two weekly challenges here is the third and final lap of the Inspire Europe Grand Prix.  Keep in mind that the participants only had 10 minutes to complete the race. 

If you have what it takes, watch for the qualification rounds of the 2017 Grand Prix to be held at Inspire 2017 (Las Vegas June 5-8, 2017)  http://www.alteryx.com/inspire-2017

[brianprestidge]

Hello Alteryx Folks - FYI - you have left the answers in the Text Input Boxes in the workflow

Brian

[TaraM]

Whoopsie! All fixed now ;) Thanks @brianprestidge

[brianprestidge]

My Solution...

Spoiler

Answers:
1) 1.817008
2) Evening (4663)
3) 2562
4) Pedestrian 1.89016
5) Male

[JohnJPS]

Spoiler

Answers:
1) 1.817008
2) Evening (4663)
3) 2562
4) Pedestrian 1.86759
5) Male

I believe a subtle difference in the numeric portion of answer (4) is due to selecting "Age of Casualty" as a numeric field; if I leave it as a V_String, I get the same answer as Brian and also those that were originally included in the problem release.  I do believe selecting it as a numeric field is "more correct" than leaving it as a string factor, though: e.g. 35 and 36 are similar ages, but "35" and "36" as string factors would have no similarity.  (So goes my thought process anyway - could be wrong!)

[brianprestidge]

I Like it John - and I reckon you're right - makes perfect sense to me!

[TaraM]

A solution has been posted.

Spoiler

[SeanAdams]

Very similar responses to those posted above.

total time across all laps: 53:33

[NicoleJohnson]

My solution, similar to the others. I'm not sure if anyone else would agree (I'm very new to Predictive analytics), but I was thinking as I did this one that it would be nice if you could take the results of the Logistic Regression and analyze them, such as being able to take all the p-values and sort/order them to be used in additional tools later on in a workflow, rather than relying on looking at the results in the browse window. Maybe there already is a way to do this, but I couldn't seem to find the right tool.

Spoiler