DETAILED ACTION
Status of Claims
Claim(s) 1-4, 6-10, 12-17 and 19-20 are pending and are examined herein.
Claim(s) 1, 8, and 14 have been Amended. Claim(s) 5, 11, and 18 are Cancelled.
Claim(s) 1-4, 6-10, 12-17 and 19-20 are rejected under 35 U.S.C. §§ 112 and101.
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 .
Response to Amendment
The amendment filed on January 20, 2026 has been entered. Claims 1-4, 6-10, 12-17 and 19-20 are pending in the application. Applicant’s amendments to the claims have been fully considered and are addressed in the rejections below.
Response to Arguments
Applicant's arguments with respect to the rejection under 35 U.S.C. § 101 filed on 01/20/2026 have been fully considered but they are not persuasive.
Applicant’s argument (Pp. 9-10 of the remarks):
Applicant argues that “Claim 1 as amended recites, inter alia, "recursively train a neural network for detecting anomalies from continuous data streams that eliminate quantile crossing when utilizing quantiles for the continuous data streams with the neural network; , . , updating a parameter in the neural network based on a pinball loss that evaluated quantiles values using the training data for each node in the tree-structure in an order designated from a root node to leaf nodes within tree-structure; .... detecting data stream anomalies from the continuous data streams with network monitoring applications that utilizes the neural network through incrementally tracked quantiles that decrease memory costs of the network monitoring applications."” and asserts that the claim is not directed to mathematical concepts or mental processes.
Examiner's response: The examiner respectfully disagrees with the applicant’s for the following reasons:
First, the claim as currently drafted exhibits a written description issue addressed in the rejection under 35 U.S.C. § 112(a). Specifically, the amend claim now explicitly recites anomaly detection and memory-reducing functionality as required, active steps of the method. This shifts the claimed invention from primarily quantile-estimation using the divide-and-conquer framework for quantile regression to one that performs anomaly detection and resource optimization, which the specification does not sufficiently describe nor does it provide examples of how to implement.
Second, the claimed invention is primarily directed to a divide-and-conquer algorithm for quantile regression, which is statistical and mathematical techniques used in statistics and econometrics to estimate conditional quantiles of a response variable. The algorithm performs a sequence of mathematical calculations and statistical analysis to estimate quantile values. This includes transforming a list of quantile values into a tree-structure by recursively dividing an interval between 0 and 1 into subintervals and estimating a relative quantile value using the lower and upper bounds of each node in the tree. This recursive divide-and-conquer approach for quantile estimation represents a statistical and mathematical process and is directed to the abstract idea of mathematical concepts and mental processes. See MPEP § 2106.04(a)(2).
Third, the neural network recitations merely represent a generic computer component that is employed as a tool to perform the abstract idea of quantile estimation. The recitation of detecting data stream anomalies in network monitoring applications merely define the field of use of quantile estimation. While quantile estimation may be used to incrementally track quantiles of a data distribution and detect anomalies. This actual process would fall within the category of mathematical concepts and mental process. Furthermore, the specification paragraph [0026] describes this limitation in the background as an intended field of use, specifically, that quantiles can be used in network monitoring applications to detect data stream anomalies, rather than as an implemented method or processas a tool to perform the quantile estimation. These recitations therefore amount to invoking a computer or other machinery in its ordinary capacity merely as a tool to perform the judicial exception, and generally linking the judicial exception into a particular technological environment or field of use. See MPEP §§ 2106.05(f) & (h).
Forth, the claim describes statistical regression analysis utilizing a divide-and-conquer approach for quantile regression, and applies a neural network as the regression estimator, rather than claiming any specific technical implementation of the neural network itself. As described in the specification, the neural network or other regression models (e.g., random forest, gradient boosting) merely serve as interchangeable estimation components. Specifically, as set forth in paragraph [0055]-[0066], the underlying quantile estimation algorithm remains unchanged while the regression estimator may be replaced with a random forest regression model or a gradient boosting regression model. This clearly establishes that the neural network is incorporated as a tool to perform the abstract idea of quantile estimation, and that the claim is not directed to any particular neural network training process. Moreover, the training is described only at a high level of generality, using a standard pinball loss function commonly used in quantile regression to evaluate quantile estimation accuracy. This pinball loss function is well-know and conventional method for calculating a loss value based on estimated quantile values, as evidenced by Steinwart et al., which states that “the so-called pinball loss for estimating conditional quantiles is well-known tool in both statistics and machine learning.” Accordingly, the claim invokes computer components in their ordinary capacity merely as a tool to perform an existing process. MPEP § 2106.05(f).
Accordingly, the claim is directed to an abstract idea and doesn’t include additional elements that are sufficient to integrate the abstract idea into a practical application.
Applicant’s Argument (Pp. 11-13 of the remarks):
Applicant further argues that “present embodiments include additional features that are more than more mathematical concepts... recursively train a neural network and update a parameter in the neural network based on a pinball loss that evaluated quantile values using the training data which is not merely directed to a mathematical concept... A human mind cannot practically "update a parameter in the neural network based on a pinball loss that evaluated quantile values using the training data" nor can a human mind practically ''train a neural network."... The present embodiments improve the functioning of machine learning models when estimating quantiles from continuous data streams... Similar to resolving the "catastrophic forgetting" issue in Desjardins, the present embodiments resolve the technical problem concerning the crossing problem when estimating quantiles with neural networks... Additionally, similar to Koninklijke... the present embodiments use a specific way to improve anomaly detection machine learning models.”
Examiner's response: The examiner respectfully disagrees with the applicant’s arguments for the following reasons:
First, with respect to Applicant’s assertion that the neural network recitations and pinball loss training are more than mere mathematical concepts and cannot be performed by the human mind, the Examiner notes that these recitations are considered as additional elements under Step 2A, Prong 2 of the Alice/Mayo framework, and are not themselves characterized as an abstract idea. However, as set forth in the earlier discussed responses above, these additional elements are not sufficient to integrate the abstract idea into a practical application. As previously discussed, the neural network is employed merely as a tool to perform the abstract idea of quantile estimation, and the training process is described only at a high level of generality using a standard pinball loss function that is well-known and conventional in quantile regression and machine learning. The mere fact that certain steps cannot be practically performed in the human mind does not, by itself, establish that the claim integrates the judicial exception into a practical application or amounts to significantly more. See MPEP § 2106.04(d).
Second, with respect to Applicant’s assertion that the present embodiments improve the functioning of machine learning models when estimating quantiles from continues data streams, the Examiner respectfully disagrees. As previously discussed, the claim is not directed to a particular neural network training method; rather, it is directed to quantile estimation using a neural network as a tool. Furthermore, the specification describes the use of quantiles in continuous data streams as an intended use of the claimed divide-and-conquer quantile estimation approach, not as a specific technical improvement to the neural network itself as applicant alleged. Additionally, it is important to note that the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. Here, the alleged improvement in resolving quantile crossing comes from the judicial exception itself (i.e., reclusively estimating quantile values), and not from any additional elements recited in the claim. The identified additional elements, specifically the neural network and the pinball loss training, are not themselves, nor in combination, amount to the alleged improvement. Rather, the judicial exception is modified with the identified additional elements at a high level of generality, without claiming any specific technical implementation that would transform the claim into a patent eligible subject matter. Accordingly, the claim does not integrate the judicial exception into a practical application. See MPEP § 2106.05(a).
Third, with respect to Applicant’s reliance on Ex parte Desjardins and Koninklijke, the Examiner respectfully disagrees that these cases are analogous to the present claims. In Desjardins, the claimed improvement was directed to specific technical solution to the problem of catastrophic forgetting with neural network training process itself, where the claim recites training steps that that provided the improvement. Here, by contrast, the present claims do not recite a specific technical implementation of the neural network training process. The neural network and the pinball loss function are recited at a high level of generality, and the specification makes clear that the neural network may be replaced with other regression models without altering the underlying quantile estimation. This demonstrates that the claimed improvement comes from the abstract divide-and-conquer quantile estimation algorithm, not in any particular neural network implementation. Similarly, in Knoinklijke where the claim recites technical method of generating check data that improved error detection systems. Here, the claim does not recite a specific technical method of improving anomaly detection systems; rather, it recites the general use of incrementally tracked quantiles in network monitoring applications as an intended filed of use of the claimed quantile estimation approach. See specification [0026]. Accordingly, neither Desjardins nor Koninklijke support Applicant’s position that the present claims integrate the abstract idea into a practical application.
In view of the above, the rejection under 35 U.S.C. § 101 is maintained.
Claim Objections
The disclosure is objected to because of the following informalities:
Applicant is advised that should claim 1 be found allowable, claim 8 will be objected to under 37 CFR 1.75 as being a substantial duplicate thereof. When two claims in an application are duplicates or else are so close in content that they both cover the same thing, despite a slight difference in wording, it is proper after allowing one claim to object to the other as being a substantial duplicate of the allowed claim. See MPEP § 608.01(m).
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claim(s) 1-4, 6-10, 12-17 and 19-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Regarding Currently amended Claim 1, the amendments introduces new functional requirements that change the scope of the claim. In particular, the claim now requires explicit anomaly detection, a recursive neural network training framework, and memory-reduction though incrementally tracked quantiles, as active steps of the method, which are not adequately supported by the original disclosure. The specification primarily describes a framework for conditional quantile estimation using neural network, rather than describing the specific operations now required by the claim. More specifically, the following amended limitations lack sufficient written description support:
The claim recites: “recursively training a neural network for detecting anomalies from continuous data streams that eliminates quantile crossing when utilizing quantiles for the continuous data streams with the neural network” lines 6-8. This limitation requires that the neural network be trained recursively to detect anomalies in continuous data streams and the link between quantile crossing elimination and its application to continuous data streams as a specific operational context for the neural network. However, the specification primarily describes a quantile regression framework for estimating conditional quantiles of a response variable distribution (see, e.g., [0001], [0003]-[0005], [0025], and [0069]-[0084]). While the specification paragraph [0026] states that in network monitoring applications “anomalies on data streams need to be detected as the data streams change dynamically over time,” but this is described as background context for why quantile tracking is useful, and not as a description of training a neural network specifically for anomaly detection from continuous data streams. The training phase described throughout the spec (e.g., paragraphs [0075]-[0079], Fig. 11) is described as training for quantile regression, not for anomaly detection from continuous data streams as an active process of the neural network training itself. For example, the specification does not describe a neural network that is recursively trained for anomaly detection, how estimated quantiles are used to identify anomalies, or any anomaly detection criteria using quantile estimation. Furthermore, the specification does not specifically describe this quantile crossing elimination in the context of utilizing quantiles for continuous data streams. The specification states that the divide-and-conquer framework ensures no crossing by restricting the range of each rx,j in [0,1], (see, paragraph [0028] and [0060]).However, this is not described in the context of utilizing quantiles for continuous data streams. Thus, the specification ties quantile crossing elimination to the general regression estimation framework, not specifically to its use with continuous data streams as claimed. Accordingly, the claim limitation introduce a new matter that lacks written description support.
Additionally, the claim recites the limitations: “recursively training a neural network ... including: transforming the list of quantile levels into a tree-structure ... dividing an interval in a range between 0 and 1 sub-intervals ... and updating a parameter in the neural network based on a pinball loss” lines 6-20. The claim as currently drafted defines the transforming and dividing steps as sub-components of the neural network training operation. However, the specification does not describe the transformation of quantile levels into a tree-structure, nor the division of the intervals into sub-intervals, as operations that occur within or as part of neural network training. The specification defines these operations as distinct and sequential. Specifically, paragraph [0050] and Fig. 5 describes a transforming component (72) and a training component (74) as separate system components. Fig. 10 further presents transformation (Block 1020) and neural network training (Block 1030) as distinct sequential steps, where transformation occurs prior to the training operation. Accordingly, the amened claim training structure is not described by the specification and therefore lacks adequate written description support.
Furthermore, the claim recites the limitation: “detecting data stream anomalies from the continuous data streams with network monitoring applications that utilizes the neural network through incrementally tracked quantiles that decrease memory costs of the network rnonitori.ng applications” lines 30-34. The specification does not provide adequate written description support for this limitation for three reasons:
First, the specification does not describe the trained neural network being used (at inference or deployed) to detect anomalies. The testing phase (Fig. 12, paragraphs [0080]-[0083]) only describes outputs quantile values, it does not define anomaly detection as an output or application of the neural network itself. Anomaly detection is described only in the background (paragraph [0026]) as a general use case for quantile tracking, not as a function performed by the trained neural network.
Second, incremental quantile tracking is referenced in paragraphs [0026]-[0027] as the intended use of the claimed divide-and-conquer framework. The specification does not describe the claimed framework as performing incremental quantile tracking as an active step or feature of the disclosed method; it only mention it as background motivation.
Third, while the specification paragraph [0027] discusses memory cost issues with prior incremental quantile methods (“a large amount of summary information must be maintained, which tends to be expensive in terms of memory”) and states that the invention addresses quantile estimation issues such as the crossing problem. However, the specification provides no disclosure that describes or demonstrate that the claimed method decreases memory const of network monitoring applications and/or does not describe any specific memory optimization technique or storage method. Further, the specification fails to provide an example how incrementally tracking quantile using the claimed process result in decreased memory consumption in network mentoring applications.
Accordingly, the claim as currently drafted introduces new functional requirements that lack adequate written description.
Regarding Currently Amended Dependent Claims 9 and 10, the claim lacks written description support for the claim limitations as currently presented. Claim 9 recites “wherein the neural network utilizes random forest regression” and claim 10 recites “wherein the neural network utilizes gradient boosting regression.” These limitations require that the neural network itself incorporates or utilizes random forest regression or gradient bossing regression within the claimed method. However, the specification (see, e.g., paragraphs [0062]-[0064] and Fig. 7) describes random forest and gradient boosting as alternative regression models that may be used instead of the neural network, rather than components or techniques that are utilized within or as part of the neural network. In particular, the disclosure indicates that the neural network may be replaced with these models, which is inconsistent with the claim’s current requirement that the neural network utilizes them.
Because the original disclosure only describes these models as replacement/substitutes for the neural network, and does not describe or suggest a neural network that utilizes random forest or gradient bossing regression, the specification does not reasonably convey a person ordinary skill in the art that the inventor was in possession of the currently presented claim feature at the time of filing. Accordingly, Claims 9 and 10 lack adequate written description support.
Regarding Currently Amended Independent Claims 8 and 14, the claims recite substantially similar limitations as corresponding to claim 1 with similar issues of written description. Thus, the same rationale applies to the independent claims 8 and 14.
Regarding Dependent Claims 2-4, 6-7, 9-10, 12-13, 15-17 and 19-20, dependent claims inherit the deficiencies of the respective parent claim.
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.
Claim(s) 9 and 10 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, for pre-AIA the applicant regards as the invention.
Regarding Currently Amended Dependent Claims 9 and 10, Claims 9 and 10 depend from parent Claim 8. Claim 8 recites the use of “neural network” throughout the claimed steps. However, Claim 9 further recites that the neural network “utilizes random forest regression,” and claim 10 recites that the neural network “utilizes gradient boosting regression.” These recitations introduces uncertainty regarding the scope of the claims. In particular, it is unclear whether the neural network includes random forest or gradient boosting regression as internal components or whether the neural network defined in claim 8 is replaced with a random forest or gradient boosting regression model.
Since the claim does not clearly define the relationship between the neural network and the recited regression utilization, one of ordinary skill in the art would not be able to determine the metes and bounds of the claimed invention. Accordingly, claims 9 and 10 recites limitations that render the claimed invention indefinite.
In view of the above, Examiner respectfully requests that Applicant thoroughly review the claims for compliance with the requirements set forth under 35 U.S.C. § 112.
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.
When considering subject matter eligibility under 35 U.S.C. 101, it must be determined whether the claim is directed to one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter (Step 1). If the claim does fall within one of the statutory categories, the second step in the analysis is to determine whether the claim is directed to a judicial exception (Step 2A). The Step 2A analysis is broken into two prongs. In the first prong (Step 2A, Prong 1), it is determined whether or not the claims recite a judicial exception (e.g., mathematical concepts, mental processes, certain methods of organizing human activity). If it is determined in Step 2A, Prong 1 that the claims recite a judicial exception, the analysis proceeds to the second prong (Step 2A, Prong 2), where it is determined whether or not the claims integrate the judicial exception into a practical application. If it is determined at step 2A, Prong 2 that the claims do not integrate the judicial exception into a practical application, the analysis proceeds to determining whether the claim is a patent-eligible application of the exception (Step 2B). If an abstract idea is present in the claim, any element or combination of elements in the claim must be sufficient to ensure that the claim integrates the judicial exception into a practical application, or else amounts to significantly more than the abstract idea itself. Applicant is advised to consult MPEP 2106 for more details of the analysis.
Under Step 1 analysis,
Claims 1-4 and 6-7 recite a method (representing a process);
Claims 8-10 and 12-13 recite a method (representing a process); and
Claims 14-17 and 19-20 recite a computer program product (representing an article of manufacture).
Therefore, each set of the claims falls into one of the four statutory categories (i.e., process, machine, article of manufacture, or composition of matter). Examiner’s Note: Regarding claims 14-17 and 19-20, the specification defines the “storge medium” to exclude signals per se, see paragraph [0117]. Therefore, claims 14-20 are within the four statutory categories of invention.
Claims 1-4, 6-10, 12-17 and 19-20
are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more, and hence is not patent-eligible subject matter.
Regarding Currently Amended Claim 1,
Step 2A Prong 1: The claim recites an abstract idea enumerated in the 2019 PEG.
transforming the list of quantile levels into a tree-structure ... until all quantile levels has been transformed to obtain relative positions of nodes of the tree-structure between an estimated lower bound value and an estimated upper bound value; dividing an interval in a range between 0 and 1 into sub-intervals by using the list of quantile levels such that each node of the tree-structure is associated with a tuple of three quantile levels; (The “transforming” step, as drafted and under its broadest reasonable interpretation, falls within the mathematical concepts and mental processes of abstract ideas. This encompasses a process/step that can be performed by humans mentally or with a pencil and paper. An individual can manually make a list of quantile/percentile levels and create/draw a tree representing different quantile levels ranges between [0, 1]. At high level of generality, the step for transforming a list of quantile levels was recited such that it can be performed by humans with the assistance of physical aids (e.g., pencil and paper), see MPEP § 2106.04(a)(2)(III). The recitation of “obtain estimated relative position” which is the result of the calculated upper and lower bound of each quantile is part of the abstract idea of mathematical concepts.)
estimating the tuple of three quantile levels, via the neural network, including a relative quantile value, a first estimated quantile value, and a second estimated quantile value, the relative quantile value as an output of the neural network, for each of the first values by using the first estimated quantile value, estimated by utilizing the output of the neural network as the estimated lower bound value, and the second estimated quantile value, estimated by utilizing the output of the neural network as the upper bound value; (The “estimating” step is directed to an abstract idea of a “mental process” and/or “mathematical concepts.” Examiner notes: the “estimating” step, as drafted, and under the broadest reasonable interpretation, covers concepts that can be performed in the human mind and/or with physical aid (e.g., pen and paper). The recitation of using generic computer component (i.e., neural network) to perform the "estimating” step does not preclude the limitation from practically being performed in the human mind and/or with physical aid (e.g., pen and paper), see MPEP § 2106.04(a)(2),(III).)
calculating the list of quantile levels based on the first estimated quantile value, the second estimated value, and the relative quantile value; (The “calculating” step is an abstract idea of “mathematical concept and “mental processes.” The “calculating” step, as drafted, and under the broadest reasonable interpretation, covers concepts that can be performed in the human mind and/or with physical aid (e.g., pen and paper). MPEP § 2106.04(a)(2)(III) indicates that the use of a physical aid (e.g., pencil and paper or a slide rule) to help perform a mental step (e.g., a mathematical calculation) does not negate the mental nature of the limitation. Thus calculating quantile values is a mental step.)
detecting data stream anomalies from the continues data streams (An abstract idea of a mental process and mathematical concepts. Examiner’s note: Under the broadest reasonable interpretation (BRI), the step “detecting data steam based on the estimated relative positions …” covers concepts that can be performed in the human mind and/or with the aid of pen and paper, see MPEP § 2106.04(a)(2)(III). Merely using quantile estimation to incrementally track quantiles and to identify anomalies represents a statistical analysis of the data distribution using regression quantile estimation. Thus, analyzing data distribution using quantile estimation encompasses mathematical concepts and mental process—concepts performed in the human mind including observation, evaluation, judgment or opinion, see MPEP § 2106.04(a)(2)(I) & (III).)
Step 2A Prong 2: Under this prong, we evaluate whether the claim recites additional elements that integrate the abstract idea into a practical application by considering the claim as a whole. The judicial exception is not integrated into a practical application.
Additional Elements Analysis:
acquiring training data represented as coordinates with first values and second values, a list of quantile levels, a lower bound of the second values, and an upper bound of the second values, wherein each of first values is a feature vector and each of the second values is a real number (Amount to no more than adding insignificant extra-solution activity to the judicial exception (e.g., mere data gathering in conjunction with the abstract idea), see MPEP § 2106.05(g).)
recursively training neural network for detecting anomalies from continuous data streams that eliminate quantile crossing when utilizing quantiles for continuous data stream with the neural network... updating a parameter in the neural network based on a pinball loss that evaluated quantile values using the training data for each node in the tree-structure in an order designated from a root node to leaf nodes within the tree-structure; (As outlined above, this limitation is subjected to 112(a) written description issue and the specification does not support the claimed neural network training for anomaly detection from continuous data streams. The recited additional elements such as training neural network for quantile estimation amounts to no more than instruction and/or using generic computer components to apply the aforementioned abstract idea. In other words, this amounts to merely reciting the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f). Additionally, the recitation of updating a parameter in the neural network based on pinball loss amounts to reciting generic model training. Examiner’s note: These recitations represents high level training of a neural network model with previously determined data. The recitation of “recursively employing a neural network” merely describes the concept of inserting a computer component (i.e., neural network) as a tool in a divide-and-conquer regression algorithm to perform relative quantile estimation. Thus, the neural network acts as an estimator to carry out the abstract idea of estimating or calculating relative quantile value. In particular, using a generic computer component or other machinery in its ordinary capacity to perform a mental/mathematical step. The recited “pinball loss” function, used to calculate loss based on the estimated quantile values, represents a standard and conventional function commonly used in the context of quantile estimation. Accordingly, the use of such a conventional loss function in combination with the generic training of a neural network amounts to nothing more than generic computer functions merely used to implement an abstract idea. As evidenced by Steinwart et al. (NPL: “Estimating conditional quantiles with the help of the pinball loss." (2011)), which states that “the so-called pinball loss for estimating conditional quantiles is a well-known tool in both statistics and machine learning.”)
The recitation of the “neural network output” amounts to mere instructions to apply the exception. In other words, the claim invokes generic computer component or other machinery in its ordinary capacity (i.e., neural network) to perform the quantile estimation (i.e., abstract idea).
detecting data stream anomalies from the continuous data streams with network monitoring applications that utilizes the neural network through incrementally tracked quantiles that decrease memory costs of the network monitoring applications. (This limitation is subjected to 112(a) written description issue, which the disclosure does not provide sufficient support of the claimed feature and the specification merely presented the concept of data stream anomalies in network monitoring applications and incremental quantile estimation as background context for why quantile tracking is useful, not as an active step performed by the neural network as claimed. Thus, this limitation merely defines the intended field of use of the claimed quantile estimation and cannot be interpreted as an active step of the claimed process. See applicant’s specifications [0026]. Accordingly, this limitation amounts to generally linking the use of a judicial exception to a particular technological environment or field of use, as discussed in MPEP § 2106.05(h). The application or use of the judicial exception in this manner does not meaningfully limit the claim by going beyond generally linking the use of the judicial exception a particular technological environment or field of use as disclosed in the Applicant’s specification.
Step 2B: Under this prong, the claim must include additional elements that amount to significantly more than the judicial exception. These elements must not be well-understood, routine, or conventional in the relevant field. When viewed individually and as an ordered combination, the claim does not include any such additional elements that are sufficient to amount to significantly more (i.e., inventive concept).
Additional Elements Analysis:
The “acquiring training data” step was considered to be extra-solution activity in Step 2A, and thus it is re-evaluated in Step 2B to determine if it is more than what is well-understood, routine, conventional activity in the field. The additional element “acquiring training data” is well-understood and conventional function because it does not amount to an inventive concept, particularly when the function is well-understood or conventional (MPEP 2106.05(g)). This appears to be “Storing and retrieving information in memory” which is well‐understood, routine, and conventional functions, as evidenced by MPEP 2106.05(d)(II)(iv).
As explained above, the recited elements in the respective claim 1 such as “training a neural network” and “updating a parameter in the neural network based on a pinball loss” to carry out the quantile estimation steps amounts to no more than mere instructions to apply the judicial exception on a computer. Thus, the high-level training of a neural network for estimating relative quantile value represents the tool that is used to apply the mental process and/or mathematical concepts. Additionally, the recited “pinball loss” function merely defines a well-known and conventional function used in the context of quantile estimation, as evidenced by Steinwart et al., which states that “the so-called pinball loss for estimating conditional quantiles is a well-known tool in both statistics and machine learning.” Mere instructions to apply an exception cannot provide an inventive concept. Accordingly, the same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
The limitation “detecting data stream anomalies from the continuous data streams with network monitoring applications that utilizes the neural network through incrementally tracked quantiles that decrease memory costs of the network monitoring applications” merely describes an intended use or field-of-use limitation. According to Applicant’s disclosure (see paragraph [0026]-[0027]), this additional element merely identifies anomaly detection as an intended application of quantile estimation within network monitoring systems, rather than reciting an implemented method or specific technical process. Thus, this limitation does not constitute an active step in the claimed process and merely links the use of the judicial exception to a particular field or use, or defines an intended result of the claimed method. Accordingly, it does not meaningfully limit the claim and cannot amount to an inventive concept.
Accordingly, when viewed as a whole, the claim is primarily directed to the abstract idea of estimating quantile values using divide-and-conquer approach, and the recited additional elements, whether considered alone or in combination with the judicial exception, do not amount to significantly more. The recited additional elements consist of generic computing and data gathering that are well-known in the field and do not add an inventive concept.
Therefore, claim 1 does not recite patent-eligible subject matter.
Regarding Original Claim 2,
Step 2A Prong 1: Claim 2, which incorporates the rejection of claim 1, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
wherein a tree recurrent neural network (RNN) is employed. (This amounts to no more than merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f). Examiner notes that the claim merely invokes a generic computer component performing a generic computer function.)
Step 2B: The claimed additional elements are not sufficient to amount to significantly more than the judicial exception.
The “RRN” merely used as a tool to perform the abstract idea. This step of computation represents a generic computer function recited at a high level of generality. Accordingly, a claim reciting a generic computer component performing a generic computer function is necessarily ineligible.
Therefore, claim 2 is ineligible.
Regarding Original Claim 3,
Step 2A Prong 1: Claim 3, which incorporates the rejection of claim 2, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
a tree long short-term memory (LSTM) is used as a cell in the tree-RNN. (This amounts to no more than merely reciting the words "apply it" (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f). Examiner notes that the claim merely invokes a generic computer component performing a generic computer function.)
Step 2B: The additional element does not amount to significantly more than the judicial exception. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 3 is ineligible.
Regarding Original Claim 4,
Step 2A Prong 1: Claim 4, which incorporates the rejection of claim 1, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
wherein the neural network is a fully connected neural network with a single output or a fully connected neural network with multiple outputs. (Amount to no more than mere instructions to apply/implement the abstract idea on a generic computer. Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f).)
Step 2B: The additional element does not amount to significantly more than the judicial exception. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 4 is ineligible.
Regarding Original Claim 6,
Step 2A Prong 1: Claim 6, which incorporates the rejection of claim 1, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
wherein the relative quantile values of the first values are displayed in a graphical format on a display of a computing device. (Amount to no more than mere instructions to apply/implement the abstract idea on a generic computer. Displaying the output of the analysis or results on a genetic computing device does not negate the abstract idea.)
Step 2B: The additional element does not amount to significantly more than the judicial exception. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 6 is ineligible.
Regarding Original Claim 7,
Step 2A Prong 1: Claim 7, which incorporates the rejection of claim 6, recite the limitation:
wherein the relative quantile values of the first values are displayed in a graphical format on a display of a computing device. (The recited illustrating “an absence of non-monotone quantile values” is part of the abstract idea of mental/mathematical concept of quantile estimation and visualization of results. It involves representing calculated numerical relationships (monotonicity of quantile) in graphical form and depicting the result of mathematical calculation.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
Displaying or visualizing the computed quantile results in a graphical format represents insignificant post-solution activity and/or presentation of the data analysis, see MPEP § 2106.05(g). This insignificant post-solution activity does not meaningfully limit the claim as only defines the results of the abstract idea.
Step 2B: The additional element does not amount to significantly more than the judicial exception.
As noted above, the addition of insignificant extra-solution activity does not amount to an inventive concept, particularly when the activity is well-understood or conventional. The “graphical format” feature merely uses generic computer display functions to show outcome of the abstract idea (absence of non-monotone quantile values). The use of a conventional graphical output to depict the result of the abstract idea is well-understood, routine, and conventional activity. This merely represents a data output in conjunction with the abstract idea. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 7 is ineligible.
Regarding Currently Amended Claim 8,
The claim recites similar limitations as corresponding claim 1. Therefore, the same analysis (subject matter eligibility analysis) that was utilized for claim 1, as described above, is equally applicable to claim 8.
Therefore, claim 8 is ineligible.
Regarding Currently Amended Claim 9,
Step 2A Prong 1: Claim 9, which incorporates the rejection of claim 8, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
wherein the neural network utilizes random forest regression. (Amount to no more than mere instructions to apply/implement the abstract idea on a generic computer. Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f).)
Step 2B: The additional element does not amount to significantly more than the judicial exception. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 9 is ineligible.
Regarding Currently Amended Claim 10,
Step 2A Prong 1: Claim 10, which incorporates the rejection of claim 8, doesn’t recite an abstract idea.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
wherein the regression model is a gradient boosting regression model. (Amount to no more than mere instructions to apply/implement the abstract idea on a generic computer. Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f).)
Step 2B: The additional element does not amount to significantly more than the judicial exception. The same analysis utilized under Step 2A Prong 2 is equally true in Step 2B.
Therefore, claim 10 is ineligible.
Regarding Previously presented Claim 12,
The claim recites similar limitations as corresponding claim 6. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 6, as described above, is equally applicable to claim 12.
Therefore, claim 12 is ineligible.
Regarding Original Claim 13,
The claim recites similar limitations as corresponding claim 7. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 7, as described above, is equally applicable to claim 13.
Therefore, claim 13 is ineligible.
Regarding Currently Amended Claim 14,
The claim recites similar limitations as corresponding claim 1. Therefore, the same analysis (subject matter eligibility analysis) that was utilized for claim 1, as described above, is equally applicable to claim 14. The only difference is that claim 1 is drawn to a method, and claim 14 is drawn to a computer program product. The recitation of “a computer program product…, comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer” merely defines computer component and instructions to implement a judicial exception, and hence the claimed additional elements listed above are merely generic elements and the implementation of the elements merely amount to no more than instruction to apply the abstract idea using a generic computer component. Therefore, the additional elements do not integrate the judicial exception into a practical application. See MPEP 2106.05(f).
Therefore, claim 14 is ineligible.
Regarding Original Claim 15,
The claim recites similar limitations as corresponding claim 2. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 2, as described above, is equally applicable to claim 15.
Therefore, claim 15 is ineligible.
Regarding Original Claim 16,
The claim recites similar limitations as corresponding claim 3. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 3, as described above, is equally applicable to claim 16.
Therefore, claim 16 is ineligible.
Regarding Original Claim 17,
The claim recites similar limitations as corresponding claim 4. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 4, as described above, is equally applicable to claim 17.
Therefore, claim 17 is ineligible.
Regarding Original Claim 19,
The claim recites similar limitations as corresponding claim 6. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 6, as described above, is equally applicable to claim 19.
Therefore, claim 19 is ineligible.
Regarding Original Claim 20,
The claim recites similar limitations as corresponding claim 7. Therefore, the same subject matter eligibility analysis (including the abstract idea) that was utilized for claim 7, as described above, is equally applicable to claim 20.
Therefore, claim 20 is ineligible.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SADIK ALSHAHARI whose telephone number is (703)756-4749. The examiner can normally be reached Monday Friday, 9 A.M - 6 P.M. ET..
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, Li Zhen can be reached on (571) 272-3768. 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.
/S.A.A./Examiner, Art Unit 2121
/Li B. Zhen/Supervisory Patent Examiner, Art Unit 2121