You can use the Google Patents Public Datasets (mirror), which can be accessed through Google BigQuery.
For example, this query (its execution is free of charge):
SELECT publication_number, assignee_harmonized.name, filing_date, country_code
-- WHERE UPPER(assignee_harmonized.name) LIKE 'ADOBE%'
Here are two links to GitHub projects that can be used to help download and parse USPTO data.
I am also in the process of building a comprehensive tool in Python that downloads, parses and builds an SQL database of all patent, application, PAIR, and classification data-sets....
The freepatentsonline.com expert search (www.freepatentsonline.com/search.html) is really robust free resource. Scroll down for the expert search.
To do the search you've described, check the box for US Patents (and uncheck any other boxes). Then enter: AN/stanford + "machine learning" The AN stands for "assignee name" and entering "machine learning" ...