EverBank and the University of Central Florida Data Mining Program are pleased to announce the 3rd Annual EverBank Cup Contest. The contest consists of three classification tasks based on data from EverBank Lending (note: all data have been obscured to hide actual balances and all personal information has been excluded).
Contest Tasks:
Contestants will be provided with one raw dataset containing three binary targets, the first target represents whether the client paid-in-full their current mortgage loan, the second target represents whether the customer filed a loan application within three months of paying off their old loan, the third target represents whether the customer eventually got the loan from the bank. The goal is to build a three stage predictive model that will accurately forecast a customer’s likelihood to pay off their current loan as the first stage model. The second stage model is to predict the likelihood of applying a new loan from EverBank within three months after paying off the old loan, and the third stage is to estimate the likelihood of all these applicants in stage to get the loan.
Contestants are required to submit the following materials for judging:
- Project report that includes
- Cover page with registration number, name and contact information. This should be the only page that identifies the contestant.
- Executive summary: one paragraph
- Data preparation: two pages
- Modeling approach and code: two pages for each modeling stage
- Results and conclusions (including test results performed on your model): two pages
- The scoring data set with the following fields:
- The ContestID
- Probability scores for the first target
- Binary decision for the first target
- Probability scores for the second target for all applicants the binary decision for the first target is “1”, i.e., “positive”
- Binary decision for the second target based on its probability score
- Probability scores for the third target for all applicants the binary decision for the second target is “1”, i.e., “positive”
- Binary decision for the third target based on its probability score
Judging:
All entries will be judged based on the following criteria:
- 40% – The contents in the Project Report. Higher weighting will be given to those providing clear analysis and reasoning behind the selected modeling approach.
- 40% – Model accuracy based on the first target:
- Accuracy Rate – 20%
- AUC (area under the ROC curve) – 10%
- Cumulative percent captured target in the top 3 deciles – 10%
- 10% – Model accuracy based on the second target’s accuracy rate
- 10% – Model accuracy based on the third target’s accuracy rate
Eligibility:
This year’s contest is open to students currently enrolled full-time or part-time at any University or College in the state of Florida. Faculty and staff are not eligible for participation. Students are required to enter the contest as an individual. Receiving assistance in this contest is grounds for elimination. By participating, contestants agree to be bound by the judges’ decisions.
How to Enter:
Click “Register Now” button and fill out the registration form at the following URL:
Once registered, a registration ID will be assigned to each contestant by the system. The contestant will receive an email with the registration ID and a link to download the datasets and data dictionary.
How to Submit Results:
On or before the due date, contestant should email the required materials (project reports and results file) to DM_Contest@everbank.com with subject line “2015 EverBank Cup Contest Results.” Include your registration ID and contact information in the email. Results file should be in this format:
Results file: SAS or CSV file with the file name being ID_x where “x” is your registration ID. The file should be compressed with ZIP. Please only return Field1, Predicted Target1 Probability, Binary (Yes/No) Decision for Target1, Predicted Target2 Probability, , Binary (Yes/No) Decision for Target2, Predicted Target3 Probability, and Binary (Yes/No) Decision for Target3 on the result dataset and do not include any other fields on the original test dataset. Please also specify your registration ID and contact information in the submission email.
Any file not meeting this specification will be rejected.
Prizes:
$2,000 in cash prizes will be awarded to the top three winners:
- First Place $1,000
- Second Place $600
- Third Place $400
Winners will be required to provide W-2 information and proof of meeting eligibility requirements before awards are presented. In addition, winners are expected to present a 15-minute overview of their methods and results at the April 15th “Big Data Symposium” at the University of Central Florida campus. Travel expenses are the responsibility of the contestant.
Important Dates: Contest start date: January 12, 2015 (data files will be available on this date)
Registration deadline: February 22, 2015
Final submissions due: On or before 11pm ET March 13, 2015
Notification of top three winners: March 31, 2015
Award ceremony: April 15, 2015
INTELLECTUAL PROPERTY
DATA
‘Data’ means the Data or Datasets linked from the Competition Website for the purpose of use by Participants in the Competition. For the avoidance of doubt, Data is deemed for the purpose of these Competition Rules to include any prototype or executable code provided to Participants by EverBank via the Website. Participants must use the Data only as permitted by these Competition Rules and any associated data use rules specified on the Competition Website.
Unless otherwise permitted by the terms of the Competition Website, Participants must use the Data solely for the purpose and duration of the Competition, including but not limited to reading and learning from the Data, analyzing the Data, modifying the Data and generally preparing your Submission and any underlying models. Participants agree to use suitable measures to prevent persons who have not formally agreed to these Competition Rules from gaining access to the Data and agree not to transmit, duplicate, publish, redistribute or otherwise provide or make available the Data to any party not participating in the Competition. Participants agree to notify EverBank immediately upon learning of any possible unauthorized transmission or unauthorized access of the Data and agree to work with EverBank to rectify any unauthorized transmission. Participants agree that participation in the Competition shall not be construed as having or being granted a license (expressly, by implication, estoppel, or otherwise) under, or any right of ownership in, any of the Data.
EXTERNAL DATA
Unless otherwise expressly stated on the Competition Website, Participants must not use data other than the Data to develop and test their models and Submissions. EverBank reserves the right in its sole discretion to disqualify any Participant who EverBank discovers has undertaken or attempted to undertake the use of data other than the Data, or who uses the Data other than as permitted according to the Competition Website and in these Competition Rules, in the course of the Competition.
——–UPDATED ON Mar-4-2015——–