Data Science

Sponsored By 10Pearls

Members Allowed : 1-4

Rules:

  1. Minimum 1 and maximum 4 members are allowed in one team.
  2. Participants can use any language and framework.
  3. Participants need to bring their own machines for the competition with all the dependencies installed for their selected language/platform.
  4. The competition will be Open Internet!
  5. Participants will be given the Training and Testing Dataset through USB or Server. (Note: Only Training Data will have target Label/Output values)
  6. Participants need to submit the following at the end of the competition:
  7. The predictions for the Testing Data in a CSV/XLS file using a USB/Server. (USB will be provided)
  8. One-page document which defines the selected approach in brief (Data cleaning methods, Data exploration methods used etc.), Feature engineering Techniques (if applied) and selected Machine Learning Model (No hyper-parameters!).

Winning Criteria:

  1. The team with Highest F1 measure will be awarded as the winner.  If multiple teams end up with the same F1 score, the decision will be made based on selected approaches.

Resource Provided:

  1. Participants will be provided with the basic stationery (sheet, pencil, pen).

Round(s):

  1. There will only be one round of about 3-5 hours (Tentative)