NON-FINAL REJECTION, FIRST DETAILED ACTION
Status of Prosecution
The present application 18/497,284, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
The application was filed in the Office on Oct. 30, 2023.
Claims 1-20 are pending and are all rejected in this rejection. Claims 1, 12 and 17 are independent claims.
Status of Claims
Claims 7 and 16 are rejected under 35 U.S.C. § 112(b) second paragraph.
Claims 1-3, 8-13 and 16-19 are rejected under 35 USC § 103 as being unpatentable over Gambietta et al. (“Gambietta”), United States Patent Application Publication 2020/0320437 published on Oct. 8, 2020, in view of non-patent literature Zhang et al. (“Zhang”), “Quantum classification algorithm with multi-class parallel training” published in 2022.
Claims 4-6, 14-15 and 20 are rejected under 35 USC § 103 as being unpatentable over Gambietta in view of Zhang and in further view of non-patent literature Hamamura et al. (“Hamamura”), “Efficient evaluation of quantum observables using entangled measurements,” published in 2020.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. § 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
Claims 7 and 16 are rejected under 35 U.S.C. § 112(b) second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Representative independent claim 1 recites in part, “a transforming component that transforms a qubit of a quantum feature map from an initial state to a transformed state based on an input value of a first classical dataset, and
a dataset component that generates a second classical dataset based on the input value and the transformed state.”
Claim 7, which depends from claim 1 recites in part: “wherein the transformed state comprises a first transformed state, wherein the qubit comprises a first qubit, wherein the quantum feature map comprises a first quantum feature map, wherein the input value comprises a first input value, wherein the dataset component generates the second classical dataset further based on a second input value and a second transformed state of a second qubit of a second quantum feature map, and wherein the second transformed state was transformed based on the second input value of the first classical dataset.”
Here, Examiner queries as to how a qubit may comprise what appears to be multiple qubits as recited by Applicant. Additionally, Examiner queries as to the reference to how the second input value used in transforming the second transformed state is “of the first classical dataset.” Further, it is unclear as to how the second qubit is transformed based on the transformed state in the parent claim as well as the second transformed state in an apparently circular fashion.
No prior art rejection is made. See MPEP 2173.06(II) (“As stated in In re Steele, 305 F.2d 859, 134 USPQ 292 (CCPA 1962), a rejection under 35 U.S.C. 103 should not be based on considerable speculation about the meaning of terms employed in a claim or assumptions that must be made as to the scope of the claims.”).
Applicant’s is invited to interview with Examiner to discuss this claim.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. § 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
-A.
Claims 1-3, 8-13 and 16-19 are rejected under 35 USC § 103 as being unpatentable over Gambietta et al. (“Gambietta”), United States Patent Application Publication 2020/0320437 published on Oct. 8, 2020, in view of non-patent literature Zhang et al. (“Zhang”), “Quantum classification algorithm with multi-class parallel training” published in 2022.
As to Claim 1, Gambietta teaches: A system comprising: a memory that stores computer-executable components; and
a processor, operatively coupled to the memory, that executes the computer-executable components stored in the memory (Gambietta: par. 0026, a system with memory executable by a processor via the memory), wherein the computer-executable components comprise:
a transforming component that transforms a qubit of a quantum feature map from an initial state to a transformed state based on an input value of a first classical dataset (Gambietta: par. 0095, “A quantum feature map implemented by quantum feature map circuit 600 functions to make input data linearly separable into categories as required by an SVM/QSVM as it imposes hyperplanes on the "lifted" (e.g., feature map applied) data”; Fig 12, par. 0113, the training data [1202] and resulting sampling of objects from the training set may be classical).
PNG
media_image1.png
743
1012
media_image1.png
Greyscale
Gambietta may not explicitly teach: a dataset component that generates a second classical dataset based on the input value and the transformed state.
Zhang teaches in general concepts related to a quantum classification algorithm that is optimized by a hybrid quantum-classical method, composed of a trainable quantum circuit and a gradient-based classical optimizer (Zhang: Abstract). Specifically, Zhang teaches that the quantum circuit provides information (second classical dataset) to a classical optimizer and iteratively operates with the quantum computer again with updates (Zhang: Fig. 1, Sec. 2, “In the training processes, the initial qubits in label register are transformed into the computational basis states by Hadamard operators, and the training data for different classes are loaded into the quantum circuit. Then, the measurements on two registers are performed for the calculation of the cost function. The parameters in cost function are optimized by a gradient-based optimization algorithm in a classical computer and re-input to the quantum circuit for next data-loading. After several iterations, the parameters are optimal when the cost function is converged”).
PNG
media_image2.png
451
535
media_image2.png
Greyscale
It would have been obvious to a person having ordinary skill in the art at a time before the effective filing date of the application to have modified the Gambietta disclosures and teachings by providing the training data to a second component for generating classical data for use as taught by Zhang. Such a person would have been motivated to do so with a reasonable expectation of success to allow for the benefits of a hybrid quantum and classical computing architecture.
As to Claim 2, Gambietta and Zhang teach the elements of claim 1.
Zhang further teaches: wherein the computer-executable components further comprise a training component that trains a classical machine learning model based on the second classical dataset (Zhang: Fig. 1, Sec. 2, “In the testing processes, the unclassified data and the optimal parameters are fed into the quantum circuit, and the classification information of the unclassified data can be obtained when the measurement results on the sample register are performed.”)
As to Claim 3, Gambietta and Zhang teach the elements of claim 2.
Zhang further teaches: wherein the input value comprises a first input value, wherein the training component further transforms the qubit of the quantum feature map from the transformed state to a further transformed state based on a second input value of the second classical dataset, and wherein the dataset component further generates a third classical dataset based on the second input value and the further transformed state (Zhang: Sec. 2, the values are iteratively fed, which is a repetition and Examiner asserts is a generating of a third classical dataset).
As to Claim 8, Gambietta and Zhang teach the elements of claim 1.
Gambietta further teaches: wherein the quantum feature map is based on the first classical dataset (Gambietta: par. 0113, the training data [1202] and resulting sampling of objects from the training set may be classical).
As to Claim 9, Gambietta and Zhang teach the elements of claim 1.
Zhang further teaches: wherein the transforming component transforms the qubit of the quantum feature map to the transformed state further based on application of a quantum gate (Gambietta: par. 0090, various quantum gates, including bit-flip, Hadamard, phase gate, etc.).
As to Claim 10, Gambietta and Zhang teach the elements of claim 1.
Gambietta further teaches: wherein the quantum gate comprises a Hadamard gate (Gambietta: par. 0090, various quantum gates, including bit-flip, Hadamard, phase gate, etc.)..
As to Claim 11, Gambietta and Zhang teach the elements of claim 9.
Gambietta further teaches: wherein the quantum gate comprises a phase gate (Gambietta: par. 0090, various quantum gates, including bit-flip, Hadamard, phase gate, etc.).
As to Claim 12, it is rejected for similar reasons as claim 1.
As to Claim 13, it is rejected for similar reasons as claim 2.
As to Claim 16, it is rejected for similar reasons as claim 7.
As to Claim 17, Gambietta teaches: A computer program product that generates a result classical dataset based on an initial classical dataset and a quantum feature map, the computer program product comprising a computer readable storage medium having program instructions embodied therewith (Gambietta: par. 0026, a system with memory executable by a processor via the memory), the program instructions executable by a processor to cause the processor to:
transform a qubit of the quantum feature map from an initial state to a transformed state based on an input value of the initial classical dataset(Gambietta: par. 0095, “A quantum feature map implemented by quantum feature map circuit 600 functions to make input data linearly separable into categories as required by an SVM/QSVM as it imposes hyperplanes on the "lifted" (e.g., feature map applied) data”; Fig 12, par. 0113, the training data [1202] and resulting sampling of objects from the training set may be classical); and
Gambietta may not explicitly teach: generate the result classical dataset based on the input value and the transformed state.
Zhang teaches in general concepts related to a quantum classification algorithm that is optimized by a hybrid quantum-classical method, composed of a trainable quantum circuit and a gradient-based classical optimizer (Zhang: Abstract). Specifically, Zhang teaches that the quantum circuit provides information (result classical dataset) to a classical optimizer and iteratively operates with the quantum computer again with updates (Zhang: Fig. 1, Sec. 2, “In the training processes, the initial qubits in label register are transformed into the computational basis states by Hadamard operators, and the training data for different classes are loaded into the quantum circuit. Then, the measurements on two registers are performed for the calculation of the cost function. The parameters in cost function are optimized by a gradient-based optimization algorithm in a classical computer and re-input to the quantum circuit for next data-loading. After several iterations, the parameters are optimal when the cost function is converged”).
It would have been obvious to a person having ordinary skill in the art at a time before the effective filing date of the application to have modified the Gambietta disclosures and teachings by providing the training data to a second component for generating classical data for use as taught by Zhang. Such a person would have been motivated to do so with a reasonable expectation of success to allow for the benefits of a hybrid quantum and classical computing architecture.
As to Claim 18, Gambietta and Zhang teach the elements of claim 17.
Zhang further teaches: further comprising program instructions to train a classical machine learning model based on the result classical dataset (Zhang: Fig. 1, Sec. 2, “In the testing processes, the unclassified data and the optimal parameters are fed into the quantum circuit, and the classification information of the unclassified data can be obtained when the measurement results on the sample register are performed.”).
As to Claim 19, Gambietta and Zhang teach the elements of claim 18.
Gambietta and Zhang as combined further teaches: further comprising program instructions to train the classical machine learning model based on the initial classical dataset (Examiner asserts that Zhang’s iterative process would necessarily use the initial classical dataset in its subsequent training).
B.
Claims 4-6, 14-15 and 20 are rejected under 35 USC § 103 as being unpatentable over Gambietta et al. (“Gambietta”), United States Patent Application Publication 2020/0320437 published on Oct. 8, 2020, in view of non-patent literature Zhang et al. (“Zhang”), “Quantum classification algorithm with multi-class parallel training” published in 2022 and in further view of non-patent literature Hamamura et al. (“Hamamura”), “Efficient evaluation of quantum observables using entangled measurements,” published in 2020.
As to Claim 4, Gambietta and Zhang teach the elements of claim 3.
Gambietta further teaches: wherein the quantum feature map comprises mappings of a first feature of the input value and a second feature of the input value (Gambietta: par. 0019, “the method includes applying, by a quantum processor, a set of quantum feature maps to the
selected objects, the set of quantum maps corresponding to a set of quantum kernels.”).
Gambietta and Zhang may not explicitly teach: wherein the transformed state comprises a first observable based on the first feature and a second observable based on the second feature.
Hamamura teaches in general concepts related to quantum-classical hybrid algorithms in noisy intermediate-scale quantum (NISQ) computers (Hamamura: Abstract). Specifically, Hamamura teaches that target observables, which are mainly Hamiltonians, and may be written as a linear combination of Pauli strings (Hamanura: Results).
It would have been obvious to a person having ordinary skill in the art at a time before the effective filing date of the application to have modified the Gambietta-Zhang disclosures and teachings by having respective observables with features as taught and suggested by Hamamura. Such a person would have been motivated to do so with a reasonable expectation of success to best perform the algorithmic computations effectively on a feature-by-feature basis.
As to Claim 5, Gambietta, Zhang and Hamamura teach the elements of claim 4.
Hamamura further teaches: wherein the transformed state comprises a stacking of the first observable and the second observable (Hamamura: p. 2, the grouping of jointly measurable Paul strings by using sequential measurements or on a tensor product basis).
As to Claim 6, Gambietta, Zhang and Hamamura teach the elements of claim 4.
Gambietta, Zhang and Hamamura further teaches: wherein the dataset component generates the second classical dataset based on: a first iteration that transforms the second classical dataset based on the first observable, and a second iteration that transforms the second classical dataset based on the second observable.
It would have been obvious to a person having ordinary skill in the art at a time before the effective filing date of the application to have further modified the Gambietta-Zhang-Hamamura disclosures and teachings by having the respective classical datasets based on their respective observables be transformed iteratively per the combination. Such a person would have been motivated to do so with a reasonable expectation of success to allow for the efficient and better expressive quality use of the hybrid classical quantum architecture.
As to Claim 14, it is rejected for similar reasons as claim 4.
As to Claim 15, it is rejected for similar reasons as claim 6.
As to Claim 20, it is rejected for similar reasons as claim 4.
Conclusion
Relevant prior art not relied upon but made of the record:
Ahmad et al. (“Ahmad”), “Quantum Machine Learning with HQC Architectures using non-Classically Simulable Feature Maps,” published in 2021 (describing challenge in selection of quantum embeddings and the use of a functionally correct quantum variational circuit).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAMES T TSAI whose telephone number is (571)270-3916. The examiner can normally be reached M-F 8-5 Eastern.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached on 571-270-5871. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000./JAMES T TSAI
/JAMES T TSAI/ Primary Examiner, Art Unit 2147
Prosecution Insights