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

[Emil_Kos]

Hi,

Fun exercise. I am not able to work with models too often.

[RashedDS]

My solution

[MatthewBr]

Finished what I believe to be correct....

Question -  4 and 5.

4) Which causality class has the lowest p-value in the model? (Pr(>|z|)) What is the coefficient? (Estimate). 

- Are you just looking at the browse tool for the lowest number? Is there another option?

5) According to the model, which gender increases the likelihood of being a fatal casualty?

- How are you determining this? 

[Daniel_Griffith]

Done!

Unable to download the Association Analysis tool due to corporate IT policies, so the last two questions are unanswered

[Daniel_Griffith]

Got access to the predictive tools, updated solution attached.

Association Analysis requires numeric info to compare against so didn't work. Looked at the solution and saw that logistic regression is what I needed.

[DanTh]

Spoiler

Managed to get through the steps with the right answers, but must say that I need to spend a fair amount of time looking at the Logistic Regression outputs to make sure I understand everything.

3/5

[ARussell34]

Good ol' regression slowed things down a bit, but I was able to find answers to all the questions!

1.) Average number of vehicles involved in an accident = 1.93

2.) Evening = the time bucket with the most accidents

3.) 2,664 = the number of accidents involving a pedestrian

4.) Logistic Regression:

     1.)  The Pedestrian Class = the class with the lowest p-value in the model at p = 0.00067***

      2.) The Male sex = the gender that increases the likelihood of being in a fatal casualty ( Estimate: 0.7537)

I don't believe I would have been fast enough in this last question to win, but I was happy to gain some more experience with the Logistic Regression tool.

[Erin]

Spoiler

First timer using regression. 

[pfaigmani]

My solution to challenge # 44. 

[LiuZhang]

Spoiler