Machine Learning FAQ
General questions for Machine Learning.
Does Machine Learning (AYX ML) support time series?
Yes, Machine Learning supports single series. Multiple series is not supported at this time.
Does Machine Learning support anomaly detection?
Not at this time. Machine Learning supports supervised binary classification, multiclass classification, and regression methods.
Does Machine Learning support unsupervised learning methods?
Not at this time. Machine Learning supports supervised binary classification, multiclass classification, and regression methods.
How many columns can my dataset have?
Machine Learning supports up to 512 columns in your dataset.
Is there a character limit for columns in my dataset?
Yes. No single column can contain more than 2000 characters.
What date-time formats does Machine Learning support?
ISO 8601 and RFC 822.
Are there any limitations on what kinds of characters my data can include?
Yes. Machine Learning doesn't support characters that are not UTF-8 encoded.
We also recommend that you remove any dollar-sign characters ($) in front of amounts so that Machine Learning can parse them as doubles or numeric values.
Is there a row limit?
We recommend that you limit your rows to less than 500,000.
Is there a limit to the size of my dataset?
Yes, the file size limit is 1.2 GB.
Where can I find out more about the AutoML library used in Machine Learning?
Machine Learning uses EvalML to build your machine learning models. To find out more about the EvalML library, go to the EvalML documentation.