Milestones

List view

• Modeling Across Time - Upgrading to a Model Capable of Temporal Updates

  We would like to train a network (likely RNN or LSTM or similar) to predict what shot will occur next in a tennis point given what has happened up to that point.

  No due date

  •2/2 issues closed
  100% complete0 open2 closed

• Tennis Agency - RL Model to choose next action

  A stretch goal for the project is to train a reinforcement learning agent to choose either the best or most likely next action (shot choice) given the previous shots in the rally / game.

  No due date

  •0/1 issues closed
  0% complete1 open0 closed

• Organization and Documentation

  Throughout the project, organization of the repo and data files (not in repo) should be maintained. Also, during/after writing code, comments / markdown should be added to aid understanding and make it easier to collaborate.

  No due date

  •1/2 issues closed
  50% complete1 open1 closed

• Regression Modeling - Predict Point Winner

  Create point-level dataset and train regression model to predict whether the server wins the point. The main goal here is to have this model serve as a foundation and baseline for future models and to figure out which features may be most informative.

  No due date

  •2/4 issues closed
  50% complete2 open2 closed

• EDA

  Initial exploration and visualization of the raw Roland Garros (mens) dataset. The goal was to understand the data's structure, completeness, and some basic statistics like the counts of forehands, backhands, serves, types of errors, etc.

  No due date

  •3/3 issues closed
  100% complete0 open3 closed