Webinar—How Data Science is Changing Football: What Every Industry Can Learn!

Q: How does DataRobot select the right blueprint for a given problem?

DataRobot doesn't "select" blueprints it creates them dynamically based on the dataset that you give it and the target you specify. Data scientists at DataRobot have spent over 7 years embedding their data science knowledge and best practices into this blueprint development process. Roughly hundreds of blueprints are created for each project. For most data science problems, you can come up with multiple viable approaches without knowing which ones will perform the best. DataRobot uses the Leaderboard to compare blueprints and identify the best ones.

Q: Can I see the effect of some of the DataRobot-created features?

DataRobot shows the effect of some of the engineered variables through the Insights tab. Both the Tree-Based Variable Importance as well as the Variable Effects sections show the engineered variables used by the models.

Q: Is it possible for users to change the preprocessing method?

Yes, you can change some of the preprocessing methods used within DataRobot on the Advanced Tuning tab of each model. If you have a custom preprocessing step that you want to include, you could do it outside of DataRobot and load in the processed dataset. This becomes seamless when using the DataRobot R/Python clients.

Q: What do asterisks on the Leaderboard metrics mean?

The asterisks indicate that the scores are computed from in-sample training data.

Q: What does a snowflake icon mean near a model in the Leaderboard?

This indicates a frozen run. A frozen run of a model is a retrained version of another model, where hyperparameters from the other model are frozen and the model is simply retrained on more observations.

Q: What does a snowflake icon mean near a model in the Leaderboard?

This indicates a frozen run. A frozen run of a model is a retrained version of another model, where hyperparameters from the other model are frozen and the model is simply retrained on more observations.

Q: Why didn't cross-validation automatically run on my dataset?

Cross-validation has a hard cutoff at 50,000 rows; if the dataset is greater than or equal to 50,000 rows, DataRobot does not run cross-validation automatically. If you require automatic cross-validation, use a dataset with fewer than 50,000 rows. If you need to get cross-validation values for a specific model, on the Leaderboard for that model click Run (under the Cross Validation column). If the dataset is larger than 800MB, DataRobot will use TVH. (Cross-validation is generally useful for smaller datasets where you would not otherwise have enough useful data using TVH.)

lhaviland · ‎07-14-2020

Football and other sports have changed dramatically by adopting Artificial Intelligence, which is certain to disrupt every business industry. Decisions across all organizations benefit greatly by injecting data science, which more accurately comprehends the complexities of real-life information. Predictions and recommendations are improved for use cases off the field (marketing, ticket pricing), in-game strategy, recruiting players, etc.

This webinar leads a lively discussion on how winning in sports translates to winning across other industries, overcoming cultural resistance, and doing analytics at scale and velocity to win the race.

Host:

Andrew Engel (DataRobot, General Manager for Sports)
Ari Kaplan (DataRobot, Director of Sports Evangelism)

Click here to watch ondemand.

If you have any questions or comments about what we've shown here, click Comment and let us know!

Webinar—How Data Science is Changing Football and More

Artificial Intelligence

AutoML

Use Cases