By Kenichi Hartman

There has been a lot of discussion regarding a need for more USPTO guidance regarding patent applications directed to machine learning. While the USPTO has not yet issued a communication dedicated to the topic of machine learning or AI, the new examples (issued January 7, 2019) provided with the The 2019 Revised Patent Subject Matter Eligibility Guidance notably includes an example claim (Example 39) directed to training a neural network.

According to the discussion provided in Example 39, the following example claim would be considered patentable with respect to USPTO Subject Matter Eligibility requirements under 35 U.S.C. 101:

A computer-implemented method of training a neural network for facial detection comprising:

  • collecting a set of digital facial images from a database;
  • applying one or more transformations to each digital facial image including mirroring, rotating, smoothing, or contrast reduction to create a modified set of digital facial images;
  • creating a first training set comprising the collected set of digital facial images, the modified set of digital facial images, and a set of digital non-facial images; training the neural network in a first stage using the first training set;
  • creating a second training set for a second stage of training comprising the first training set and digital non-facial images that are incorrectly detected as facial images after the first stage of training; and
  • training the neural network in a second stage using the second training set.

Remarkably, this claim is determined to NOT be directed to a judicial exception at all under Alice Step 2A, Prong 1 of the updated Subject Matter Eligibility Guidelines. As a result, further analysis of the claim under Alice Step 2A, Prong 2, or Alice Step 2B is deemed unnecessary.

The reasoning given for this determination is that the claim does not recite any of the various types of judicial exceptions enumerated in the new 2019 PEG, for example:

  1. The claim does not recite any mathematical relationships, formulas, or calculations. While some of the limitations may be based on mathematical concepts, the mathematical concepts are not recited in the claims.
  2. The claim does not recite a mental process because the steps are not practically performed in the human mind.
  3. The claim does not recite any method of organizing human activity such as a fundamental economic concept or managing interactions between people.

The claim given in Example 39 was clearly designed to reflect claims directed to training methods for machine learning that are currently being prosecuted at the USPTO. The fact that such a claim is now unambiguously considered patent-eligible is great news for the AI/Machine Learning community.