DETAILED ACTION
This action is in response to the application filed 01/30/2024. Claims 1-20 are pending and have been examined.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Specification
The disclosure is objected to because of the following informalities:
[0047, 0068, 0071, 0074, 0079, 0082, 0086, 0098, 00113]: The mathematical formulas are fuzzy and difficult to parse.
Appropriate correction is required.
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.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 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.
Claim 1 recites “updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples” in its third limitation. It’s unclear whether the “feature-weights for the original examples and the new examples” consists of weights already present in the original and / or new examples, or of new weights intended to be associated with the original and / or new examples. Thus, the scope of the claim is rendered indefinite. This deficiency is present in substantially similar independent claims 8 and 15, and inherited by all dependent claims. This is interpreted as updating the set of feature-weights, initially calculated for the original examples, with feature-weights calculated incorporating the original and new examples.
Claim 5 recites “updating, by the classical computer program, the correlation ratio for the original examples with the Pearson correlation for the original examples and the new examples” in its second limitation. There is insufficient antecedent basis for “the Pearson correlation” in the claim. It’s additionally unclear whether the Pearson coefficient is for the new examples in addition to the original examples or not. This deficiency is present in substantially similar claims 12 and 19. “the Pearson correlation” is interpreted as “a Pearson correlation” for the original examples and not necessarily for the new examples.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed inventions are directed to non-statutory subject matter without significantly more.
Claim 1
Step 1: The claim recites “A method”, and is therefore directed to the statutory category of process
Step 2A Prong 1: The claim recites the following judicial exception(s)
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: This can be performed as a mental process. One can mentally decide feature-weights for the data examples.
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: This can be performed as a mental process. One can mentally update the feature-weights.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
receiving, by a classical computer program, a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwriting, by the classical computer program, values stored in classical memory based on the updated feature-weights: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
loading, by the classical computer program, the feature-weights for the dataset and new data into a first quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
loading, by the classical computer program, the overwritten values stored in classical memory into a second quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
for a number of decision trees, instructing, by the classical computer program, a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: This is an instance of using a random forest to calculate labels, a well-known technique, and thus insignificant extra-solution activity (MPEP 2106.05(g)).
receiving, by the classical computer program, the labels from the quantum computer: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
receiving, by a classical computer program, a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwriting, by the classical computer program, values stored in classical memory based on the updated feature-weights: This is an instance of storing information in memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
loading, by the classical computer program, the feature-weights for the dataset and new data into a first quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
loading, by the classical computer program, the overwritten values stored in classical memory into a second quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
for a number of decision trees, instructing, by the classical computer program, a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: This is an instance of using a random forest to calculate labels, a well-known technique, as noted by Trost (Method And Apparatus For Detecting Malicious Websites, published 7/10/2014, US 20140196144 A1): “ Random forest classification is a known method in the field of statistics whereby a classification is made of an input based upon an existing dataset.” (Trost, [0020])
receiving, by the classical computer program, the labels from the quantum computer: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
Claim 2
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein the feature-weight calculation method comprises calculation of a Pearson correlation: A Pearson correlation calculation is directed to a mathematical concept.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 3
Step 1: The claim recites a process, as in claim 2
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein a method for calculating the Pearson correlation comprises:
calculating, by the classical computer program, the Pearson correlation for features in the original examples: A Pearson correlation calculation is directed to a mathematical concept.
updating, by the classical computer program, the Pearson correlation for the original examples with the Pearson correlation for the original examples and the new examples: A Pearson correlation calculation is directed to a mathematical concept.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 4
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein the feature-weight calculation method comprises calculation of a correlation ratio: A correlation ratio calculation is directed to a mathematical concept.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 5
Step 1: The claim recites a process, as in claim 4
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein a method for calculating the correlation ratio comprises:
calculating, by the classical computer program, the correlation ratio for features in the original examples: A correlation ratio calculation is directed to a mathematical concept.
updating, by the classical computer program, the correlation ratio for the original examples with the Pearson correlation for the original examples and the new examples: A correlation ratio calculation is directed to a mathematical concept.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 6
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein the feature-weight calculation method is based on a task for the dataset: Feature-weight calculation can still be performed as a mental process. One can mentally decide on feature weights based on some intended task for the dataset.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s)
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
Claim 7
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites no further judicial exception(s)
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
wherein the quantum-supervised clustering creates a specified number of clusters: Executing quantum-supervised clustering is still mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
wherein the quantum-supervised clustering creates a specified number of clusters: Executing quantum-supervised clustering is still mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
Claim 8
Step 1: The claim recites “A system”, and is therefore directed to the statutory category of machine
Step 2A Prong 1: The claim recites the following judicial exception(s)
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: This can be performed as a mental process. One can mentally decide feature-weights for the data examples.
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: This can be performed as a mental process. One can mentally update the feature-weights.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
a quantum computer: This is mere instruction to execute judicial exceptions with a generic quantum computer (MPEP 2106.05(f)).
a classical computer comprising a computer processor and executing a classical computer program: This is mere instruction to execute judicial exceptions with a generic classical computer (MPEP 2106.05(f)).
wherein the classical computer program is configured to
receive a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
calculate feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
update the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwrite values stored in classical memory based on the updated feature-weights: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
load the feature-weights for the dataset and new data into a first quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
load the overwritten values stored in classical memory into a second quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
for a number of decision trees, instruct a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: This is an instance of using a random forest to calculate labels, a well-known technique, and thus insignificant extra-solution activity (MPEP 2106.05(g)).
receive the labels from the quantum computer: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
a quantum computer: This is mere instruction to execute judicial exceptions with a generic quantum computer (MPEP 2106.05(f)).
a classical computer comprising a computer processor and executing a classical computer program: This is mere instruction to execute judicial exceptions with a generic classical computer (MPEP 2106.05(f)).
wherein the classical computer program is configured to
receive a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
calculate feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
update the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwrite values stored in classical memory based on the updated feature-weights: This is an instance of storing information in memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
load the feature-weights for the dataset and new data into a first quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
load the overwritten values stored in classical memory into a second quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
for a number of decision trees, instruct a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: : This is an instance of using a random forest to calculate labels, a well-known technique, as noted by Trost (Method And Apparatus For Detecting Malicious Websites, published 7/10/2014, US 20140196144 A1): “ Random forest classification is a known method in the field of statistics whereby a classification is made of an input based upon an existing dataset.” (Trost, [0020])
receive the labels from the quantum computer: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
Claims 9-14
Step 1: Claims 9-14 recite a machine, as in claim 8.
Step 2A Prong 1: Claims 9-14 recite the same judicial exception(s) as claims 2-7, respectively.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through any additional elements. The analysis of claims 9-14 at this step mirrors that of claims 2-7, respectively, with the exception that claims 9-14 are directed to “a classical computer comprising a computer processor and executing a classical computer program”, said program performing operations mirroring those of claims 2-7. This is a mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s). The analysis of claims 9-14 at this step mirrors that of claims 2-7, with the exception that claims 9-14 are directed to “a classical computer comprising a computer processor and executing a classical computer program”, said program performing operations mirroring those of claims 2-7. This is mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Claim 15
Step 1: The claim recites “A non-transitory computer readable storage medium”, and is therefore directed to the statutory category of article of manufacture
Step 2A Prong 1: The claim recites the following judicial exception(s)
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: This can be performed as a mental process. One can mentally decide feature-weights for the data examples.
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: This can be performed as a mental process. One can mentally update the feature-weights.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps: This is mere instruction to perform the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
receiving a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
calculating feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
updating the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwriting values stored in classical memory based on the updated feature-weights: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
loading the feature-weights for the dataset and new data into a first quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
loading the overwritten values stored in classical memory into a second quantum-accessible data structure: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
for a number of decision trees, instructing a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: This is an instance of using a random forest to calculate labels, a well-known technique, and thus insignificant extra-solution activity (MPEP 2106.05(g)).
receiving the labels from the quantum computer: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps: This is mere instruction to perform the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
receiving a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
calculating feature-weights for the original examples using the feature-weight calculation method: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
updating the feature-weights for the original examples with feature-weights for the original examples and the new examples: This is mere instruction to execute a judicial exception with generic computer hardware (MPEP 2106.05(f)).
overwriting values stored in classical memory based on the updated feature-weights: This is an instance of storing information in memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
loading the feature-weights for the dataset and new data into a first quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
loading the overwritten values stored in classical memory into a second quantum-accessible data structure: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
for a number of decision trees, instructing a quantum computer
to query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
to execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data: This is mere instruction to perform quantum-supervised clustering based on a judicial exception in a generic manner (MPEP 2106.5(f)).
to grow a depth for the tree and to store a centroid at each depth: This is mere instruction to grow a decision tree in a generic manner (MPEP 2106.05(f)).
to calculate labels for a regression task and/or a classification task: This is an instance of using a random forest to calculate labels, a well-known technique, as noted by Trost (Method And Apparatus For Detecting Malicious Websites, published 7/10/2014, US 20140196144 A1): “ Random forest classification is a known method in the field of statistics whereby a classification is made of an input based upon an existing dataset.” (Trost, [0020])
receiving the labels from the quantum computer: This is an instance of receiving data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.)
Claims 16-20
Step 1: Claims 16-20 recite an article of manufacture, as in claim 15.
Step 2A Prong 1: Claims 16-20 recite the same judicial exception(s) as claims 2-6, respectively.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through any additional elements. The analysis of claims 16-20 at this step mirrors that of claims 2-6, respectively, with the exception that claims 16-20 are directed to “A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps”, said steps comprising operations mirroring those of claims 2-6. This is a mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s). The analysis of claims 16-20 at this step mirrors that of claims 2-6, with the exception that claims 16-20 are directed to “A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps”, said steps comprising operations mirroring those of claims 2-6. This is mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
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, 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 6-8, 13-15, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Liu et al. (Feature-Weighting and Clustering Random Forest, International Journal of Computational Intelligence Systems Vol. 14(1), 2021, pp. 257–265), hereafter referred to as Liu, in view of Dahlgren et al. (METHOD AND SYSTEM FOR VALIDATING FINANCIAL EVENTS AND SECURITY EVENTS, published 2/2/2023, US 20230032963 A1), hereafter referred to as Dahlgren, and further in view of Decheng et al. (Improved quantum clustering analysis based on the weighted distance and its application, Heliyon, Volume 4, Issue 11, November 2018), hereafter referred to as Decheng.
Regarding claim 1, Liu discloses [a] method, comprising:
receiving, by a classical computer program, a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“In UCI [28], we select a benchmark of 31 datasets (plurality of original examples) for experiments. From the perspective of features, these datasets include numeric, categorical and mixed data. The number of prediction features ranges from 3 to 240. The number of the samples ranges from 101 to 20000, and the number of class labels (plurality of parameters) ranges from 2 to 26. All data have no missing values.” (Liu, page 262, left column, paragraph 4)
“In this paper, we use the Relief-F algorithm (feature-weight calculation method) [24,25] to weight features. The algorithm picks m samples. For each sample R, its knn nearest neighbors are searched in each class” (Liu, page 259, right column, paragraph 7)
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method:
PNG
media_image1.png
355
489
media_image1.png
Greyscale
(Liu, page 260, right column, Algorithm 1)
“In step 3 of Algorithm 1, Relief-F is used to get the weights” (Liu, page 260, left column, paragraph 6). Feature weights are calculated on D, a subset of the original examples.
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples; overwriting, by the classical computer program, values stored in classical memory based on the updated feature-weights; loading, by the classical computer program, the feature-weights for the dataset and new data into a first quantum-accessible data structure; loading, by the classical computer program, the overwritten values stored in classical memory into a second quantum-accessible data structure:
PNG
media_image2.png
392
487
media_image2.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 8, the set of feature-weights is updated for each node as the tree grows. Weights are saved (in some form of data structure). By virtue of how computer memory works, this is overwriting values stored in classical memory.
for a number of decision trees, instructing, by the classical computer program, a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement:
“RF is proposed by Breiman in [5]. In that paper, bootstrap sampling was used on training set (examples) to obtain a large number of random sampling subsets, and a DT was constructed based on each sampling subset” (Liu, page 257, left column, paragraph 3). The structure holding the training data is a data structure.
“RF and Bagging use bootstrap method to independently and randomly select the subsets of the raw training samples with replacement” (Liu, page 261, right column, paragraph 2). To randomly query the data structure holding the training data, its state (size of table, values of elements, indices) must be queried.
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
“In this paper, a novel method of node split of the decision trees is proposed, which adopts feature-weighting and clustering. This method can combine multiple numerical features, multiple categorical features or multiple mixed features. Based on the framework of RF, we use this split method to construct decision trees” (Liu, page 257, Abstract)
“The forest (number of decision trees) construction process is shown in Algorithm 3” (Liu, page 262, left column, paragraph 1)
PNG
media_image3.png
285
588
media_image3.png
Greyscale
(Liu, page 262, left column, Algorithm 3)
“According to (15), we use the samples in Do to estimate the voting confidence degree 𝛼 of each leaf node in the corresponding tree
PNG
media_image4.png
32
567
media_image4.png
Greyscale
In (15), the acc and err represent the number of correctly classified samples and misclassified samples of Do on the leaf node, respectively.” (Liu, page 261, right column, paragraph 2). Voting confidence is calculated with an error measure between the predicted label of the inferred data point (the label of the leaf it was directed to) and the true labeled value. This system is an example of supervised learning.
grow a depth for the tree and to store a centroid at each depth:
“The number of classes of the current node is regarded as the clustering number k. The n samples in training set D will be divided into k disjoint subsets, C1, C2, … , Ck. Firstly, class centroids are treated as the initial centers of k clusters, respectively, represented by 𝝁1, 𝝁2, ..., 𝝁k.” (Liu, page 259, right column, paragraph 3)
“In FWCRF, the trees are constructed on the basis of the top-down and recursion, and the split should be stopped, only when all samples in the current node are the same class or the feature subset used cannot distinguish these samples. The specific process is shown in Algorithm 2.” (Liu, page 261, right column, paragraph 1)
PNG
media_image5.png
472
586
media_image5.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 10, the tree is gradually grown to some depth as it recursively calls ‘grow’ on nodes to produce child nodes. As seen in line 8, class centroids are saved for each node (at each depth).
calculate labels for a regression task and/or a classification task:
“We use the DTs generated by this method as the base classifiers, and construct the ensemble classifier (FWCRF) based on the framework of RF” (Liu, page 265, left column, paragraph 1)
“the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
receiving, by the classical computer program, the labels from the quantum computer: “the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
Liu relates to random forests constructed with feature weighted-clustering and is analogous to the claimed invention.
While Liu fails to disclose the further limitations of the claim, Dahlgren discloses
receiving, by a classical computer program, a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“The data (original examples) obtained for the model 154a, the data ultimately being normalized for the model 154a is obtained, for example, from questionnaires or inquiries in response to the platform 150” (Dahlgren, [0095])
“The calculating of the weights is continuous, as new data points (new examples) enter the validation engine 154, for example, during training, as training is an iterative and cumulative process” (Dahlgren, [0081])
“The AI model 154a of the engine 154, as trained, for example, receives feedback (plurality of parameters) from the human investigations team so the machine learning component can learn whether the recommendation to contact the user and the type of the contact, was in line with the human team's recommendations” (Dahlgren, [0094])
calculating, by the classical computer program, feature-weights for the original examples using the feature-weight calculation method: “The supervised tuning at block 306b typically also involves weighting features resulting from data points” (Dahlgren, [0124])
updating, by the classical computer program, the feature-weights for the original examples with feature-weights for the original examples and the new examples: “the weighting for features may be adjusted and as additional data points (new examples) are added to the cluster, new data features, based on new data points, may be added, for example, new data point being the number of dogs the walker has as clients. This new feature, as well as additional new features, are then typically weighted with respect to the existing features” (Dahlgren, [0126])
Dahlgren relates to supervised clustering and random forest formation and is analogous to the claimed invention. Liu teaches a method of creating random forests based on feature weights of a dataset. The claimed invention improves upon this method by updating feature weights with new data. Dahlgren teaches a method of continually updating feature weights with new data, applicable to Liu. A person of ordinary skill in the art would have recognized that updating Liu’s model with new data would lead to the predictable result of training the random forest model with more data, and would improve the known device by making the training data set more robust and indicative of population parameters, leading to better models (MPEP 2143 I. (D) Applying a known technique to a known device (method, or product) ready for improvement to yield predictable results).
While Dahlgren fails to disclose the further limitations of the claim, Decheng discloses [a] method, comprising:
loading, by the classical computer program, the feature-weights for the dataset and new data into a first quantum-accessible data structure; loading, by the classical computer program, the overwritten values stored in classical memory into a second quantum-accessible data structure:
“In this paper, quantum clustering using the weighted distance by the Fuzzy-ANP method is studied. This paper consists of two parts. First, we have improved the cluster method by introducing a new weighing distance in the quantum cluster.” (Decheng, page 17, paragraph 3)
“weight W reflects the importance of each property (feature)” (Decheng, page 6, paragraph 1)
“To determine the effect of an element on its standard, the weight vector of the obtained elements must be combined to construct a supermatrix (quantum-accessible data structure)” (Decheng, page 8, paragraph 1). By the nature of how computer memory works, memory is overwritten to store the supermatrix of weights.
for a number of decision trees, instructing, by the classical computer program, a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: “the quantum clustering is the process in which the distribution of particles (quantum state) is estimated on the basis of their potential function when the wave function is given” (Decheng, page 3, paragraph 3); “Therefore, the distribution of the particles is finally determined by the potential energy function (estimated quantum state)” (Decheng, page 4, paragraph 1)
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
PNG
media_image7.png
735
659
media_image7.png
Greyscale
(Decheng, page 6, paragraph 2)
Decheng relates to quantum weighted clustering methods and is analogous to the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the existing combination to use quantum clustering in lieu of traditional clustering, as disclosed by Decheng. Decheng’s quantum clustering model has numerous benefits over traditional classical computing techniques, including the ability to cluster around arbitrary shapes, not requiring data preprocessing. It also outperforms many similar contemporary methods on randomized real data. See Decheng, page 17, paragraph 3.
Regarding claim 6, the rejection of claim 1 in view of Liu, Dahlgren, and Decheng is incorporated. Liu further discloses a method, wherein the feature-weight calculation method is based on a task for the dataset: “we estimate the correlations between features and label to weight features. When calculating the distance from a sample to a cluster center, we give a larger weight to the feature strongly related to the label (classification task label) that enlarges the contribution of the feature to the distance. Otherwise, we give a smaller weight that reduces the contribution of the uncorrelated feature to the distance.” (Liu, page 259, right column, paragraph 6)
Regarding claim 7, the rejection of claim 1 in view of Liu, Dahlgren, and Decheng is incorporated. Liu further discloses a method, wherein the quantum-supervised clustering creates a specified number of clusters: “The number of classes of the current node is regarded as the clustering number k (number of clusters)” (Liu, page 259, right column, paragraph 3)
Regarding claim 8, Liu discloses a classical computer comprising a computer processor and executing a classical computer program, wherein the classical computer program is configured to:
receive a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“In UCI [28], we select a benchmark of 31 datasets (plurality of original examples) for experiments. From the perspective of features, these datasets include numeric, categorical and mixed data. The number of prediction features ranges from 3 to 240. The number of the samples ranges from 101 to 20000, and the number of class labels (plurality of parameters) ranges from 2 to 26. All data have no missing values.” (Liu, page 262, left column, paragraph 4)
“In this paper, we use the Relief-F algorithm (feature-weight calculation method) [24,25] to weight features. The algorithm picks m samples. For each sample R, its knn nearest neighbors are searched in each class” (Liu, page 259, right column, paragraph 7)
calculate feature-weights for the original examples using the feature-weight calculation method:
PNG
media_image1.png
355
489
media_image1.png
Greyscale
(Liu, page 260, right column, Algorithm 1)
“In step 3 of Algorithm 1, Relief-F is used to get the weights” (Liu, page 260, left column, paragraph 6). Feature weights are calculated on D, a subset of the original examples.
update the feature-weights for the original examples with feature-weights for the original examples and the new examples; overwrite values stored in classical memory based on the updated feature-weights; loade the feature-weights for the dataset and new data into a first quantum-accessible data structure; loade the overwritten values stored in classical memory into a second quantum-accessible data structure:
PNG
media_image2.png
392
487
media_image2.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 8, the set of feature-weights is updated for each node as the tree grows. Weights are saved (in some form of data structure). By virtue of how computer memory works, this is overwriting values stored in classical memory.
for a number of decision trees, instructe a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement:
“RF is proposed by Breiman in [5]. In that paper, bootstrap sampling was used on training set (examples) to obtain a large number of random sampling subsets, and a DT was constructed based on each sampling subset” (Liu, page 257, left column, paragraph 3). The structure holding the training data is a data structure.
“RF and Bagging use bootstrap method to independently and randomly select the subsets of the raw training samples with replacement” (Liu, page 261, right column, paragraph 2). To randomly query the data structure holding the training data, its state (size of table, values of elements, indices) must be queried.
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
“In this paper, a novel method of node split of the decision trees is proposed, which adopts feature-weighting and clustering. This method can combine multiple numerical features, multiple categorical features or multiple mixed features. Based on the framework of RF, we use this split method to construct decision trees” (Liu, page 257, Abstract)
“The forest (number of decision trees) construction process is shown in Algorithm 3” (Liu, page 262, left column, paragraph 1)
PNG
media_image3.png
285
588
media_image3.png
Greyscale
(Liu, page 262, left column, Algorithm 3)
“According to (15), we use the samples in Do to estimate the voting confidence degree 𝛼 of each leaf node in the corresponding tree
PNG
media_image4.png
32
567
media_image4.png
Greyscale
In (15), the acc and err represent the number of correctly classified samples and misclassified samples of Do on the leaf node, respectively.” (Liu, page 261, right column, paragraph 2). Voting confidence is calculated with an error measure between the predicted label of the inferred data point (the label of the leaf it was directed to) and the true labeled value. This system is an example of supervised learning.
grow a depth for the tree and to store a centroid at each depth:
“The number of classes of the current node is regarded as the clustering number k. The n samples in training set D will be divided into k disjoint subsets, C1, C2, … , Ck. Firstly, class centroids are treated as the initial centers of k clusters, respectively, represented by 𝝁1, 𝝁2, ..., 𝝁k.” (Liu, page 259, right column, paragraph 3)
“In FWCRF, the trees are constructed on the basis of the top-down and recursion, and the split should be stopped, only when all samples in the current node are the same class or the feature subset used cannot distinguish these samples. The specific process is shown in Algorithm 2.” (Liu, page 261, right column, paragraph 1)
PNG
media_image5.png
472
586
media_image5.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 10, the tree is gradually grown to some depth as it recursively calls ‘grow’ on nodes to produce child nodes. As seen in line 8, class centroids are saved for each node (at each depth).
calculate labels for a regression task and/or a classification task:
“We use the DTs generated by this method as the base classifiers, and construct the ensemble classifier (FWCRF) based on the framework of RF” (Liu, page 265, left column, paragraph 1)
“the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
receive the labels from the quantum computer: “the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
Liu relates to random forests constructed with feature weighted-clustering and is analogous to the claimed invention.
While Liu fails to disclose the further limitations of the claim, Dahlgren discloses a classical computer comprising a computer processor and executing a classical computer program: “The computer system 120, which includes the platform 150, includes components, such as various processors, storage/memory, modules, engines, models, interfaces and the like. As used herein, a "module", for example, includes a component for storing instructions (e.g., machine readable instructions) for performing one or more processes, and including or associated with processors, e.g., the CPU 140, for executing the instructions.” (Dalhgren, [0025])
… wherein the classical computer program is configured to
receive a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“The data (original examples) obtained for the model 154a, the data ultimately being normalized for the model 154a is obtained, for example, from questionnaires or inquiries in response to the platform 150” (Dahlgren, [0095])
“The calculating of the weights is continuous, as new data points (new examples) enter the validation engine 154, for example, during training, as training is an iterative and cumulative process” (Dahlgren, [0081])
“The AI model 154a of the engine 154, as trained, for example, receives feedback (plurality of parameters) from the human investigations team so the machine learning component can learn whether the recommendation to contact the user and the type of the contact, was in line with the human team's recommendations” (Dahlgren, [0094])
calculate feature-weights for the original examples using the feature-weight calculation method: “The supervised tuning at block 306b typically also involves weighting features resulting from data points” (Dahlgren, [0124])
update the feature-weights for the original examples with feature-weights for the original examples and the new examples: “the weighting for features may be adjusted and as additional data points (new examples) are added to the cluster, new data features, based on new data points, may be added, for example, new data point being the number of dogs the walker has as clients. This new feature, as well as additional new features, are then typically weighted with respect to the existing features” (Dahlgren, [0126])
Dahlgren relates to supervised clustering and random forest formation and is analogous to the claimed invention. Liu teaches a method of creating random forests based on feature weights of a dataset. The claimed invention improves upon this method by updating feature weights with new data. Dahlgren teaches a method of continually updating feature weights with new data, applicable to Liu. A person of ordinary skill in the art would have recognized that updating Liu’s model with new data would lead to the predictable result of training the random forest model with more data, and would improve the known device by making the training data set more robust and indicative of population parameters, leading to better models (MPEP 2143 I. (D) Applying a known technique to a known device (method, or product) ready for improvement to yield predictable results).
While Dahlgren fails to disclose the further limitations of the claim, Decheng discloses a quantum computer, able to:
load the feature-weights for the dataset and new data into a first quantum-accessible data structure; load the overwritten values stored in classical memory into a second quantum-accessible data structure:
“In this paper, quantum clustering using the weighted distance by the Fuzzy-ANP method is studied. This paper consists of two parts. First, we have improved the cluster method by introducing a new weighing distance in the quantum cluster.” (Decheng, page 17, paragraph 3)
“weight W reflects the importance of each property (feature)” (Decheng, page 6, paragraph 1)
“To determine the effect of an element on its standard, the weight vector of the obtained elements must be combined to construct a supermatrix (quantum-accessible data structure)” (Decheng, page 8, paragraph 1). By the nature of how computer memory works, memory is overwritten to store the supermatrix of weights.
for a number of decision trees, instruct a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: “the quantum clustering is the process in which the distribution of particles (quantum state) is estimated on the basis of their potential function when the wave function is given” (Decheng, page 3, paragraph 3); “Therefore, the distribution of the particles is finally determined by the potential energy function (estimated quantum state)” (Decheng, page 4, paragraph 1)
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
PNG
media_image7.png
735
659
media_image7.png
Greyscale
(Decheng, page 6, paragraph 2)
Decheng relates to quantum weighted clustering methods and is analogous to the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the existing combination to use quantum clustering in lieu of traditional clustering, as disclosed by Decheng. Decheng’s quantum clustering model has numerous benefits over traditional classical computing techniques, including the ability to cluster around arbitrary shapes, not requiring data preprocessing. It also outperforms many similar contemporary methods on randomized real data. See Decheng, page 17, paragraph 3.
The analysis of claims 13-14 mirrors that of claims 6-7, with the exception that claims 13-14 are directed to generic computer hardware which executes the methods of claims 6-7. This generic hardware is taught by Dahlgren, as discussed regarding claim 8. Thus, claims 13-14 are rejected under the same rationales used for claims 6-7, respectively.
Regarding claim 15, Liu discloses steps comprising:
receiving a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“In UCI [28], we select a benchmark of 31 datasets (plurality of original examples) for experiments. From the perspective of features, these datasets include numeric, categorical and mixed data. The number of prediction features ranges from 3 to 240. The number of the samples ranges from 101 to 20000, and the number of class labels (plurality of parameters) ranges from 2 to 26. All data have no missing values.” (Liu, page 262, left column, paragraph 4)
“In this paper, we use the Relief-F algorithm (feature-weight calculation method) [24,25] to weight features. The algorithm picks m samples. For each sample R, its knn nearest neighbors are searched in each class” (Liu, page 259, right column, paragraph 7)
calculating feature-weights for the original examples using the feature-weight calculation method:
PNG
media_image1.png
355
489
media_image1.png
Greyscale
(Liu, page 260, right column, Algorithm 1)
“In step 3 of Algorithm 1, Relief-F is used to get the weights” (Liu, page 260, left column, paragraph 6). Feature weights are calculated on D, a subset of the original examples.
updating the feature-weights for the original examples with feature-weights for the original examples and the new examples; overwriting values stored in classical memory based on the updated feature-weights; loading the feature-weights for the dataset and new data into a first quantum-accessible data structure; loading the overwritten values stored in classical memory into a second quantum-accessible data structure:
PNG
media_image2.png
392
487
media_image2.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 8, the set of feature-weights is updated for each node as the tree grows. Weights are saved (in some form of data structure). By virtue of how computer memory works, this is overwriting values stored in classical memory.
for a number of decision trees, instructing a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement:
“RF is proposed by Breiman in [5]. In that paper, bootstrap sampling was used on training set (examples) to obtain a large number of random sampling subsets, and a DT was constructed based on each sampling subset” (Liu, page 257, left column, paragraph 3). The structure holding the training data is a data structure.
“RF and Bagging use bootstrap method to independently and randomly select the subsets of the raw training samples with replacement” (Liu, page 261, right column, paragraph 2). To randomly query the data structure holding the training data, its state (size of table, values of elements, indices) must be queried.
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
“In this paper, a novel method of node split of the decision trees is proposed, which adopts feature-weighting and clustering. This method can combine multiple numerical features, multiple categorical features or multiple mixed features. Based on the framework of RF, we use this split method to construct decision trees” (Liu, page 257, Abstract)
“The forest (number of decision trees) construction process is shown in Algorithm 3” (Liu, page 262, left column, paragraph 1)
PNG
media_image3.png
285
588
media_image3.png
Greyscale
(Liu, page 262, left column, Algorithm 3)
“According to (15), we use the samples in Do to estimate the voting confidence degree 𝛼 of each leaf node in the corresponding tree
PNG
media_image4.png
32
567
media_image4.png
Greyscale
In (15), the acc and err represent the number of correctly classified samples and misclassified samples of Do on the leaf node, respectively.” (Liu, page 261, right column, paragraph 2). Voting confidence is calculated with an error measure between the predicted label of the inferred data point (the label of the leaf it was directed to) and the true labeled value. This system is an example of supervised learning.
grow a depth for the tree and to store a centroid at each depth:
“The number of classes of the current node is regarded as the clustering number k. The n samples in training set D will be divided into k disjoint subsets, C1, C2, … , Ck. Firstly, class centroids are treated as the initial centers of k clusters, respectively, represented by 𝝁1, 𝝁2, ..., 𝝁k.” (Liu, page 259, right column, paragraph 3)
“In FWCRF, the trees are constructed on the basis of the top-down and recursion, and the split should be stopped, only when all samples in the current node are the same class or the feature subset used cannot distinguish these samples. The specific process is shown in Algorithm 2.” (Liu, page 261, right column, paragraph 1)
PNG
media_image5.png
472
586
media_image5.png
Greyscale
(Liu, page 261, right column, Algorithm 2). As seen in line 10, the tree is gradually grown to some depth as it recursively calls ‘grow’ on nodes to produce child nodes. As seen in line 8, class centroids are saved for each node (at each depth).
calculate labels for a regression task and/or a classification task:
“We use the DTs generated by this method as the base classifiers, and construct the ensemble classifier (FWCRF) based on the framework of RF” (Liu, page 265, left column, paragraph 1)
“the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
receiving the labels from the quantum computer: “the
l
a
b
e
l
i
is calculated by (3).
PNG
media_image6.png
57
541
media_image6.png
Greyscale
” (Liu, page 259, right column, paragraph 3)
Liu relates to random forests constructed with feature weighted-clustering and is analogous to the claimed invention.
While Liu fails to disclose the further limitations of the claim, Dahlgren discloses A non-transitory computer readable storage medium, including instructions stored thereon, which when read and executed by one or more computer processors, cause the one or more computer processors to perform steps: “In an exemplary embodiment of the disclosure, one or more tasks according to exemplary embodiments of method and/or system as described herein are performed by a data processor, such as a computing platform for executing a plurality of instructions. Optionally, the data processor includes a volatile memory for storing instructions and/or data and/or a non-volatile storage, for example, non-transitory storage media such as a magnetic hard-disk and/or removable media, for storing instructions and/or data” (Dahlgren, [0170])
Said steps comprising:
receiving a dataset comprising a plurality of original examples, a plurality of new examples, a plurality of parameters, and a selection of a feature-weight calculation method:
“The data (original examples) obtained for the model 154a, the data ultimately being normalized for the model 154a is obtained, for example, from questionnaires or inquiries in response to the platform 150” (Dahlgren, [0095])
“The calculating of the weights is continuous, as new data points (new examples) enter the validation engine 154, for example, during training, as training is an iterative and cumulative process” (Dahlgren, [0081])
“The AI model 154a of the engine 154, as trained, for example, receives feedback (plurality of parameters) from the human investigations team so the machine learning component can learn whether the recommendation to contact the user and the type of the contact, was in line with the human team's recommendations” (Dahlgren, [0094])
calculating feature-weights for the original examples using the feature-weight calculation method: “The supervised tuning at block 306b typically also involves weighting features resulting from data points” (Dahlgren, [0124])
updating the feature-weights for the original examples with feature-weights for the original examples and the new examples: “the weighting for features may be adjusted and as additional data points (new examples) are added to the cluster, new data features, based on new data points, may be added, for example, new data point being the number of dogs the walker has as clients. This new feature, as well as additional new features, are then typically weighted with respect to the existing features” (Dahlgren, [0126])
Dahlgren relates to supervised clustering and random forest formation and is analogous to the claimed invention. Liu teaches a method of creating random forests based on feature weights of a dataset. The claimed invention improves upon this method by updating feature weights with new data. Dahlgren teaches a method of continually updating feature weights with new data, applicable to Liu. A person of ordinary skill in the art would have recognized that updating Liu’s model with new data would lead to the predictable result of training the random forest model with more data, and would improve the known device by making the training data set more robust and indicative of population parameters, leading to better models (MPEP 2143 I. (D) Applying a known technique to a known device (method, or product) ready for improvement to yield predictable results).
While Dahlgren fails to disclose the further limitations of the claim, Decheng discloses [a] method, comprising:
loading the feature-weights for the dataset and new data into a first quantum-accessible data structure; loading the overwritten values stored in classical memory into a second quantum-accessible data structure:
“In this paper, quantum clustering using the weighted distance by the Fuzzy-ANP method is studied. This paper consists of two parts. First, we have improved the cluster method by introducing a new weighing distance in the quantum cluster.” (Decheng, page 17, paragraph 3)
“weight W reflects the importance of each property (feature)” (Decheng, page 6, paragraph 1)
“To determine the effect of an element on its standard, the weight vector of the obtained elements must be combined to construct a supermatrix (quantum-accessible data structure)” (Decheng, page 8, paragraph 1). By the nature of how computer memory works, memory is overwritten to store the supermatrix of weights.
for a number of decision trees, instructing a quantum computer to
query quantum states for the first quantum-accessible data structure and the second quantum-accessible data structure for new examples using random sampling with replacement: “the quantum clustering is the process in which the distribution of particles (quantum state) is estimated on the basis of their potential function when the wave function is given” (Decheng, page 3, paragraph 3); “Therefore, the distribution of the particles is finally determined by the potential energy function (estimated quantum state)” (Decheng, page 4, paragraph 1)
execute quantum-supervised clustering with the quantum states and the feature-weights for the dataset and new data:
PNG
media_image7.png
735
659
media_image7.png
Greyscale
(Decheng, page 6, paragraph 2)
Decheng relates to quantum weighted clustering methods and is analogous to the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the existing combination to use quantum clustering in lieu of traditional clustering, as disclosed by Decheng. Decheng’s quantum clustering model has numerous benefits over traditional classical computing techniques, including the ability to cluster around arbitrary shapes, not requiring data preprocessing. It also outperforms many similar contemporary methods on randomized real data. See Decheng, page 17, paragraph 3.
The analysis of claim 20 mirrors that of claim 6, with the exception that claim 20 is directed to generic computer hardware which executes the methods of claims 6. This generic hardware is taught by Dahlgren, as discussed regarding claim 15. Thus, claim 20 is rejected under the same rationales used for claim 6.
Claims 2-3, 9-10, and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Liu et al. (Feature-Weighting and Clustering Random Forest, International Journal of Computational Intelligence Systems Vol. 14(1), 2021, pp. 257–265), hereafter referred to as Liu, in view of Dahlgren et al. (METHOD AND SYSTEM FOR VALIDATING FINANCIAL EVENTS AND SECURITY EVENTS, published 2/2/2023, US 20230032963 A1), hereafter referred to as Dahlgren, and further in view of Decheng et al. (Improved quantum clustering analysis based on the weighted distance and its application, Heliyon, Volume 4, Issue 11, November 2018), hereafter referred to as Decheng, and Monaghan et al. (PREDICTIVE DATA ANALYSIS USING CUSTOM-PARAMETERIZED DIMENSIONALITY REDUCTION, published 7/29/2021, US 20210232954 A1), hereafter referred to as Monaghan.
Regarding claim 2, the rejection of claim 1 in view of Liu, Dahlgren, and Decheng is incorporated. While Liu, Dahlgren, and Decheng fail to disclose the further limitations of the claim, Kudo discloses a method, wherein the feature-weight calculation method comprises calculation of a Pearson correlation: “the predictive inference computing entity 106 determines an association value (feature-weight) (e.g., a statistical association value) for the predictive input feature and the target feature. The association value may describe any measure of association between the predictive input feature and the target feature. Examples of the noted association measures for a predictive input feature and a target feature include … a Pearson correlation coefficient ratio for the predictive input feature and the target feature” (Monaghan, [0082]).
Monaghan relates to feature significance in machine learning models and is analogous to the claimed invention. The existing combination teaches a method of clustering data based on feature weights. The claimed invention differs from this method by calculating Pearson correlations for feature weights. Monaghan teaches a method of teaching Pearson correlations for feature weights. Because both the existing combination and Monaghan teach the use of metrics for measuring feature significance, it would have been obvious to a person of ordinary skill in the art to substitute the existing combination’s metrics for calculating feature significance for the Pearson correlation to achieve the predictable result of measuring feature significance based on statistical correlation with output (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
Regarding claim 3, the rejection of claim 2 in view of Liu, Dahlgren, Decheng, and Monaghan is incorporated. Monaghan, in combination with Dahlgren, discloses a method, wherein a method for calculating the Pearson correlation comprises:
calculating, by the classical computer program, the Pearson correlation for features in the original examples: (Dahlgren) “The supervised tuning at block 306b typically also involves weighting features resulting from data points” (Dahlgren, [0124]). Combined with Monaghan’s Pearson correlation calculation of features, Pearson correlations on the data points are calculated.
updating, by the classical computer program, the Pearson correlation for the original examples with the Pearson correlation for the original examples and the new examples: “the weighting for features may be adjusted and as additional data points (new examples) are added to the cluster, new data features, based on new data points, may be added, for example, new data point being the number of dogs the walker has as clients. This new feature, as well as additional new features, are then typically weighted with respect to the existing features” (Dahlgren, [0126]). Combined with Monaghan’s Pearson correlation calculation of features, updated Pearson correlations are calculated with the new data.
Monaghan relates to feature significance in machine learning models and is analogous to the claimed invention. The existing combination teaches a method of clustering data based on feature weights. The claimed invention differs from this method by calculating Pearson correlations for feature weights. Monaghan teaches a method of teaching Pearson correlations for feature weights. Because both the existing combination and Monaghan teach the use of metrics for measuring feature significance, it would have been obvious to a person of ordinary skill in the art to substitute the existing combination’s metrics for calculating feature significance for the Pearson correlation to achieve the predictable result of measuring feature significance based on statistical correlation with output (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
The analysis of claims 9-10 mirrors that of claims 2-3, with the exception that claims 9-10 are directed to generic computer hardware which executes the methods of claims 2-3. This generic hardware is taught by Dahlgren, as discussed regarding claim 8. Thus, claims 9-10 are rejected under the same rationales used for claims 2-3, respectively.
The analysis of claims 16-17 mirrors that of claims 2-3, with the exception that claims 16-17 are directed to generic computer hardware which executes the methods of claims 2-3. This generic hardware is taught by Dahlgren, as discussed regarding claim 15. Thus, claims 16-17 are rejected under the same rationales used for claims 2-3, respectively.
Claims 4-5, 11-12, and 18-19 are rejected under 35 U.S.C. 103 as being unpatentable over Liu et al. (Feature-Weighting and Clustering Random Forest, International Journal of Computational Intelligence Systems Vol. 14(1), 2021, pp. 257–265), hereafter referred to as Liu, in view of Dahlgren et al. (METHOD AND SYSTEM FOR VALIDATING FINANCIAL EVENTS AND SECURITY EVENTS, published 2/2/2023, US 20230032963 A1), hereafter referred to as Dahlgren, and further in view of Decheng et al. (Improved quantum clustering analysis based on the weighted distance and its application, Heliyon, Volume 4, Issue 11, November 2018), hereafter referred to as Decheng.
Regarding claim 4, the rejection of claim 1 in view of Liu, Dahlgren, and Decheng is incorporated. While the aforementioned references fail to disclose the further limitations of the claim, Deng discloses a method, wherein the feature-weight calculation method comprises calculation of a correlation ratio: “There may be many correlation ratios (feature-weight[s]) that are generated for each combination of member target characteristic and job posting feature” (Deng, [0021]).
Deng relates to feature significance in machine learning models and is analogous to the claimed invention. The existing combination teaches a method of clustering data based on feature weights. The claimed invention differs from this method by calculating correlation ratios for feature weights. Deng teaches a method of teaching correlation ratios for feature weights. Because both the existing combination and Deng teach the use of metrics for measuring feature significance, it would have been obvious to a person of ordinary skill in the art to substitute the existing combination’s metrics for calculating feature significance for the correlation ratio to achieve the predictable result of measuring feature significance based on statistical correlation with output (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
Regarding claim 5, the rejection of claim 4 in view of Liu, Dahlgren, Decheng, and Deng is incorporated. Deng, in combination with Dahlgren, discloses a method, wherein a method for calculating the correlation ratio comprises:
calculating, by the classical computer program, the correlation ratio for features in the original examples: (Dahlgren) “The supervised tuning at block 306b typically also involves weighting features resulting from data points” (Dahlgren, [0124]). Combined with Deng’s correlation ratio calculation of features, correlation ratios on the data points are calculated.
updating, by the classical computer program, the correlation ratio for the original examples with the Pearson correlation for the original examples and the new examples: “the weighting for features may be adjusted and as additional data points (new examples) are added to the cluster, new data features, based on new data points, may be added, for example, new data point being the number of dogs the walker has as clients. This new feature, as well as additional new features, are then typically weighted with respect to the existing features” (Dahlgren, [0126]). Combined with Deng’s correlation ratio calculation of features, updated correlation ratios are calculated with the new data.
Deng relates to feature significance in machine learning models and is analogous to the claimed invention. The existing combination teaches a method of clustering data based on feature weights. The claimed invention differs from this method by calculating correlation ratios for feature weights. Deng teaches a method of teaching correlation ratios for feature weights. Because both the existing combination and Deng teach the use of metrics for measuring feature significance, it would have been obvious to a person of ordinary skill in the art to substitute the existing combination’s metrics for calculating feature significance for the correlation ratio to achieve the predictable result of measuring feature significance based on statistical correlation with output (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
The analysis of claims 11-12 mirrors that of claims 4-5, with the exception that claims 11-12 are directed to generic computer hardware which executes the methods of claims 4-5. This generic hardware is taught by Dahlgren, as discussed regarding claim 8. Thus, claims 11-12 are rejected under the same rationales used for claims 4-5, respectively.
The analysis of claims 18-19 mirrors that of claims 4-5, with the exception that claims 18-19 are directed to generic computer hardware which executes the methods of claims 4-5. This generic hardware is taught by Dahlgren, as discussed regarding claim 15. Thus, claims 18-19 are rejected under the same rationales used for claims 4-5, respectively.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Khadiev et al. (The Quantum Version Of Classification Decision Tree Constructing Algorithm C5.0, published 2019, arXiv:1907.06840v1) discloses a method of constructing a decision tree using quantum computing.
Liu et al. (Semi-Supervised Self-Training Feature Weighted Clustering Decision Tree and Random Forest, IEEE Access, vol. 8, pp. 128337-128348, 2020) discloses a method of performing semi-supervised learning with weighted cluster-based decision trees.
Srinivasan et al. (VEGETATION MANAGEMENT USING PREDICTED OUTAGES, filed 10/2/2023, US 2025/0111449 A1) discloses a method of performing random sampling with replacement on a quantum computer.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Aaron P Gormley whose telephone number is (571)272-1372. The examiner can normally be reached Monday - Friday 12:00 PM - 8:00 PM EST.
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, Michelle T Bechtold can be reached at (571) 431-0762. 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.
/AG/Examiner, Art Unit 2148 /MICHELLE T BECHTOLD/Supervisory Patent Examiner, Art Unit 2148
Prosecution Insights