Predictive Analytics Competition
Allstate Insurance Company will be sponsoring the second annual predictive analytics competition
in conjunction with the conference. The competition is open to all students registered for
the conference. Students may compete in teams consisting of at most 4 members. Monetary
prizes will be awarded to the top three teams: first place receives a $500 gift card, second place a
$300 gift card, and third place a $200 gift card. The top three finishers will be determined by
Allstate representatives and notified prior to the conference. The top three finishers will be
given a time slot at the conference in order to informally present their modeling approach.
For those with little modeling experience, Allstate has constructed a few Quick Start
Guides (available on the competition site linked below) that introduce logistic regression
and data manipulation in the R programming language. However, you are not required to use the R
programming language during the competition. A brief description of the problem
is provided below:
Description
The city of Chicago is required to inspect eating establishments on a regular basis to ensure
safety and health standards are met. Given the large number of establishments in the city, and
the shortage in qualified food inspectors, the city must use a data-driven approach to identify
establishments most at risk to fail inspections.
There are many factors, from location to past records to license type, which can be analyzed to
predict the likelihood of a failed inspection. Using data from the city of Chicago, the goal of
this competition is to predict the likelihood of an establishment failing a food inspection.
Asking Questions During the Competition
The competition is being hosted on
Kaggle in Class; there is a public forum on this site
to ask questions. Allstate representatives will be checking this forum daily, and should
be able to respond to questions within 2 days.
How to Participate
The competition is now open and will run through 5:59pm on Tuesday, April 19, 2016. You may begin
participating at any point prior to the competition's closing. A full description of the
competition and the rules for participation can be found on
Kaggle in Class, the site hosting the competition. In order to participate, perform
the following steps:
- Click on the following link:
kaggle.com/join/RH_food_inspection
- If you do not currently have a Kaggle account, you will be asked to create one.
Note: you must use a university email account to access this competition.
- After joining, if you are redirected to the Kaggle homepage, simply click the above link
again, and you will be directed to the competition site.
- If you are unable to join the competition site, please email the conference administrators.
We hope you will consider participating!
|