Automated Machine Learning: Past, Present and Future

preview_player
Показать описание
Since their first introduction in Version 10 of the Wolfram Language, machine learning superfunctions like Predict and Classify have tried to bridge the gap between beginner and specialist users by offering a fully automated yet customizable pipeline. Over time, new tools like feature extraction, distribution fitting and anomaly detection have been added. This talk will review the current set of our machine learning capabilities and look at what might come in the future.
Рекомендации по теме
Комментарии
Автор

All well and good what you show here Giulio.

But rather super boring compared to the
big MLs in the scene.

How about it Giulio, if you at Wolfram would
finally start to initiate your own Mashine
Learning project?

One that can show the capacity of the NN in
a mind blowing way.

You ask for an example of what that could be?

Well the Wolfram Community will not have
a hard time to give you suggestions ...

silberlinie