Skip to content

Machine Learning experiment presented at Microsoft Tech Summit 2016. Using Microsoft Azure Machine Learning Studio we'll clean the historic data, and use it to predict next season stats for the NBA players.

gk-8/Predict-NBA-Stats

Repository files navigation

NBA Stats prediction

Machine Learning experiment presented at Microsoft Tech Summit 2016. Using Microsoft Azure Machine Learning Studio we'll clean the historic data, and use it to predict next season stats for the NBA players.

Training Experiment

We'll take advantage of Azure Machine Learning Studio capabilities to transform, clean, structure and analyze our data in order to get the best results.

Training Schema (https://historicnbastats.blob.core.windows.net/images/training_schema.PNG)

Training Experiment 1(https://historicnbastats.blob.core.windows.net/images/training_exp1.PNG) Training Experiment 2(https://historicnbastats.blob.core.windows.net/images/training_exp2.PNG)

Predictive Experiment

And then we can predict 2017 stats and use them from everywhere we want (for example, a WebAPI-based conversational Bot solution):

Predictive Schema (https://historicnbastats.blob.core.windows.net/images/predictive_schema.PNG)

Predictive Experiment(https://historicnbastats.blob.core.windows.net/images/predictive_experiment.PNG)


References

[2]: Blog post with similar experiment using NodeJS and Python... http://fabianbuentello.io/blog/NBA_Machine_Learning_Tutorial/

About

Machine Learning experiment presented at Microsoft Tech Summit 2016. Using Microsoft Azure Machine Learning Studio we'll clean the historic data, and use it to predict next season stats for the NBA players.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages