The keys to the success of QML patents

In this co-published paper, Laura Compton of Haseltine Lake Kempner takes a hands-on look at how to formulate claims and applications for quantum machine learning inventions to meet EPO eligibility requirements.

In quantum machine learning (QML), classical machine learning algorithms, or expensive subroutines thereof, are typically adapted to run on a quantum computer. QML uses quantum resources to improve execution time and/or performance of classic machine learning algorithms.

Aspects of QML that may be patentable include using a quantum computer to more efficiently perform all or part of a classical machine learning algorithm (for example, using a quantum computer to more efficiently calculate classical distances for nearest neighbor, kernel and clustering methods), or to run a model yourself (for example, reformulating a stochastic model as a quantum system). Other related aspects include reformulating an optimization problem so that it can be solved using a quantum computer.

Another aspect of QML that may be patentable includes improvements to existing QML algorithms or models (for example, an improvement that reduces the depth of the quantum circuit needed to run the algorithm or model, and/or uses gates that are less complex, and/or avoids repetition of certain subroutines of the algorithm). Some improvements may be specific to the problem being solved (for example, modifying the operations applied to a quantum computing device so that a more limited space of possible solutions to an optimization problem is then searched by the device).

Inventions related to these aspects will be considered as patentable subject matter at the EPO when the quantum computing device is an integral part of the invention.

For such inventions, the independent claims will likely make some reference to the quantum computing device and the manner in which the algorithm is adapted to be implemented on it. The dependent claims, if not the independent claim itself, must:

  1. specify the initial state of the qubits of the quantum computing device;
  2. the variables that this initial state represents;
  3. how the qubits are manipulated according to the algorithm;
  4. the output obtained by measurement; and
  5. which represents the output.

Given the “technical” requirements of the EPO, it is recommended to have a dependent claim specifying how the output of the quantum computing device and/or the output of the machine learning model is then used in a technical process.

When the invention relates to more general QML methods, or improvements to such methods (which can be applied to a wide variety of problems in a wide range of fields), it is also recommended to provide a number of different use cases showing how the invention can be applied to various practical problems in the dependent claims or the specification.

Quantum computing in general, as well as QML, is a rapidly evolving and complex field. As such, preparing applications that meet the adequacy and clarity requirements of the EPO can be challenging. Therefore, it is best practice when preparing patent specifications to include a full mathematical description of the quantum implementation of the algorithm or model, in addition to how each operation applied to the qubits relates to the algorithm or model being implemented (e.g. describe how a series of operations applied to the qubits are representative of an objective function to be minimized).

For inventions related to improving existing QML algorithms or models, a detailed description of how the changes in the quantum circuit enable the improvement should be included. As with any rapidly evolving field where there is a lack of universally accepted terminologies, for applications relating to quantum computing in general, the terms used in the claims of the application must be defined in the specification.

Finally, experimental data may be particularly useful in terms of demonstrating an improvement in speed or accuracy over the prior art and may be useful in aiding inventive step in subsequent prosecution. It is also worth considering the technical problem the invention solves in terms of why classical processes have drawbacks that make them commercially or technically unviable (eg too slow for real-time implementation).

In summary, the above points can be used to assist in the preparation of QML inventions suitable for filing with the EPO and can be used to give the applicant the best possible chance of obtaining a commercially useful patent.

Laura Compton is a patent attorney in the Bristol offices of Haseltine Kempnermeer

Previous articles by Haseltine Lake Kempner authors in this series can be found here:

How to secure AI patents in Europe?

Drafting AI Patent Applications for Success at the EPO – Eligibility and Claim Formulation

Drafting AI patent applications for success at the EPO – drafting the full specification

Technology Trends – Why Patent Your Hidden AI?

Google and Samsung top the list of AI-related patent applicants at the EPO

The EPO and UKIPO Approaches to AI and Patentable Matter

How Revised EPO Guidelines Affect Treatment of AI Inventions

Monetizing data, machine learning’s most valuable asset

Leave a Comment

Your email address will not be published. Required fields are marked *