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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 10/06/2025 has been entered.
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. 17441761, filed on 03/27/2019.
Response to Arguments
Applicant’s argument filed 10/06/2025 have been fully considered but they are not persuasive.
Applicant’s Argument: On page 6 of Applicant’s response, applicant states “The subject matter recited in claim 1 amounts to “significantly more.” In particular, as described in paragraph 0051 of the specification, for example, “since the outlier detection device 1 continues to update a weight vector while acquiring an observed signal one by one, it has an advantage that it does not need to input and learn an observed signal in advance.””
Examiner’s Response: Applicant’s argument is not persuasive. An important consideration in determining whether a claim improves technology is the extent to which the claim covers a particular solution to a problem or a particular way to achieve a desired outcome, as opposed to merely claiming the idea of a solution or outcome (see MPEP 2106.05(a)). The amended claims do not provide sufficient details to describe any technological improvement. If the specifications explicitly set forth an improvement but in a conclusory manner (see MPEP 2106.04(d)(1): a bare assertion of an improvement without the detail necessary to be apparent to a person of ordinary skill in the art), the examiner should not determine the claim improves technology.
During examination, the examiner should analyze the "improvements" consideration by evaluating the specification and the claims to ensure that a technical explanation of the asserted improvement is present in the specification, and that the claim reflects the asserted improvement (see MPEP §2106.05(a)). The MPEP (§2106.05(a)(II)) also warns, “it is important to keep in mind that an improvement in the abstract idea itself (e.g. a recited fundamental economic concept) is not an improvement in technology.” Here, the alleged improvement in the form of “update a weight vector” is an improvement to the abstract idea of a mental process that can be performed in the human mind.
The additional element of a “sensor” is mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f).
Applicant’s Argument: On pages 7-8 of Applicant’s response, applicant states “Although Caritu discloses detecting a spike in error signal, as pointed out in the Office Action, such disclosure only relates to detecting whether the moving average of the norms exceeds a threshold, but does not disclose whether the number of such exceeding is over a predetermined number of times.
It is particularly noted that Caritu does not disclose anything relating to “a predetermined number of times.””
Examiner’s Response: Applicant’s argument is not persuasive. The Oostendorp reference has been included to explicitly teach “a predetermined number of times.”.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitations uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are:
“a learning unit configured to” in claims 1, 3, and 6
“a norm calculation unit configured to”, “a determination unit configured to” in claims 1, 2, and 7
“a reservoir computing step of”, “a learning step of”, “a norm calculation step of”, and “a determination step of” in claim 4
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation) recite sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
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 1-7 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Claim 1:
Subject Matter Eligibility Analysis Step 1:
Claim 1 recites “A system comprising” and is thus a machine, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
“” (a mathematical calculation)
“, calculate an error between the inner product and the observed signal, and update the weight vector using a value obtained by applying an adaptive filter to the error” (a mathematical calculation)
“” (a mathematical calculation)
“generate a time series data of the norm” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. evaluation)
“in a case where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” (a mental process, i.e. judgement)
Claim 1 therefore recites an abstract idea.
Subject Matter Eligibility Analysis Step 2A Prong 2:
“an outlier detection device; a sensor; and an observed signal receiver that acquires an observed signal from the sensor, the outlier detection device comprising” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a reservoir computer which includes an input layer, a reservoir main unit including a plurality of neurons connected to each other by synapses, and a read-out that is configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning unit configured to acquire the observed signal from the observed signal receiver, ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a norm calculation unit configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a determination unit configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 1 is directed to the abstract idea.
Subject Matter Eligibility Analysis Step 2B:
“an outlier detection device; a sensor; and an observed signal receiver that acquires an observed signal from the sensor, the outlier detection device comprising” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a reservoir computer which includes an input layer, a reservoir main unit including a plurality of neurons connected to each other by synapses, and a read-out that is configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning unit configured to acquire the observed signal from the observed signal receiver, ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a norm calculation unit configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a determination unit configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
The additional elements as disclosed above alone or in combination do not recite significantly more than the abstract idea itself as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 1 is subject-matter ineligible.
Regarding Claim 2:
Subject Matter Eligibility Analysis Step 2A Prong 1:
“” (a mental process that can be performed in the human mind, i.e. judgement)
“” (a mathematical calculation)
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the determination unit is configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“the norm calculation unit is configured to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
Regarding Claim 4:
Subject Matter Eligibility Analysis Step 1:
Claim 4 recites “An outlier detection method comprising” and is thus a machine, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
“a reservoir computing step of outputting, , an inner product of a weight vector and an activity value vector, each element of which is an activity value output by each of a plurality of neurons of a reservoir main unit of the reservoir computer,
“, calculating an error between the inner product and the observed signal, and updating the weight vector using a value obtained by applying an adaptive filter to the error” (a mathematical calculation)
“a norm calculation step of sequentially calculating a norm of the weight vector updated in the learning step” (a mathematical calculation)
“generating a time series data of the norm” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. evaluation)
“a determination step of determining whether an outlier is included in the observed signal on a basis of at least one of the norms calculated in the norm calculation step, in a case where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” (a mental process, i.e. judgement)
Claim 4 therefore recites an abstract idea.
Subject Matter Eligibility Analysis Step 2A Prong 2:
“ by a read-out of a reservoir computer of an outlier detection device in a system, of a reservoir main unit of the reservoir computer, the plurality of neurons being connected to each other by synapses of the reservoir main unit on a basis of an input to an input layer of the reservoir computer” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning step of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, ” (This step is directed to data gathering, which is understood to be insignificant extra solution activity - see MPEP 2106.05(g))
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 4 is directed to the abstract idea.
Subject Matter Eligibility Analysis Step 2B:
“ by a read-out of a reservoir computer of an outlier detection device in a system, of a reservoir main unit of the reservoir computer, the plurality of neurons being connected to each other by synapses of the reservoir main unit on a basis of an input to an input layer of the reservoir computer” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning step of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, ” (This step is directed to transmitting or receiving information, which is understood to be insignificant extra solution activity and well understood, routine and conventional activity of transmitting and receiving data as identified by the court - see MPEP 2106.05(d))
The additional elements as disclosed above alone or in combination do not recite significantly more than the abstract idea itself as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 4 is subject-matter ineligible.
Regarding Claim 5:
Subject Matter Eligibility Analysis Step 1:
Claim 5 recites “A non-transitory storage medium storing an outlier detection program causing a computer to execute”, and is thus a machine, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
“ in an outlier detection device of a system, ” (a mathematical calculation)
“ , calculating an error between the inner product and the observed signal, and updating the weight vector using a value obtained by applying an adaptive filter to the error” (a mathematical calculation)
“a norm calculation function of sequentially calculating a norm of the weight vector updated in the learning function” (a mathematical calculation)
“generating a time series data of the norm” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. judgement)
“a determination function of determining whether an outlier is included in the observed signal on a basis of at least one of the norms calculated in the norm calculation function, in a case where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” (a mental process, i.e. judgement)
Claim 5 therefore recites an abstract idea.
Subject Matter Eligibility Analysis Step 2A Prong 2:
“a reservoir computing function of having an input layer ...” (merely specifies a particular technological environment in which the abstract idea is to take place, ie. a field of use, and thus does not integrate the abstract idea into a practical application nor cannot provide significantly more than the abstract idea itself - see MPEP 2106.05(h))
“a reservoir main unit including a plurality of neurons connected to each other by synapses, and a read-out for outputting in an outlier detection device of a system, each element of which is an activity value output from each of the plurality of neurons on a basis of an input to the input layer” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning function of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, ” (This step is directed to data gathering, which is understood to be insignificant extra solution activity - see MPEP 2106.05(g))
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 5 is directed to the abstract idea.
Subject Matter Eligibility Analysis Step 2B:
“a reservoir computing function of having an input layer ...” (merely specifies a particular technological environment in which the abstract idea is to take place, ie. a field of use, and thus does not integrate the abstract idea into a practical application nor cannot provide significantly more than the abstract idea itself - see MPEP 2106.05(h))
“a reservoir main unit including a plurality of neurons connected to each other by synapses, and a read-out for outputting in an outlier detection device of a system, each element of which is an activity value output from each of the plurality of neurons on a basis of an input to the input layer” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“a learning function of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, ” (This step is directed to transmitting or receiving information, which is understood to be insignificant extra solution activity and well understood, routine and conventional activity of transmitting and receiving data as identified by the court - see MPEP 2106.05(d))
The additional elements as disclosed above alone or in combination do not recite significantly more than the abstract idea itself as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 5 is subject-matter ineligible.
Regarding Claims 3 and 6:
Subject Matter Eligibility Analysis Step 2A Prong 1:
“” (a mathematical calculation)
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the learning unit is configured to use a systolic array when the adaptive filter is applied to ” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
Regarding Claim 7:
Subject Matter Eligibility Analysis Step 2A Prong 1:
“” (a mental process that can be performed in the human mind, i.e. judgement)
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the determination unit is configured to ” (mere instructions to apply the exception using a generic computer component - see 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.
Claims 1, 4-5, and 7 are rejected under 35 U.S.C. 103 as being unpatentable by Rao (US10162378B1) in view of Oostendorp (US20210349455A1).
Regarding claim 1, Rao teaches:
“A system comprising: an outlier detection device; a sensor; and an observed signal receiver that acquires an observed signal from the sensor, the outlier detection device comprising” ([abstract, col. 12, lines 31-56, Figure 6], Rao describes a neuromorphic processor (system) for signal denoising and separation. Figure 6 shows the architecture of the signal denoising module (outlier detection device), which contains antennae (sensor) and ADC frontends (observed signal receiver). Figure 6 shows the frontends receives data from the antennae.)
“a reservoir computer which includes an input layer” ([col. 10, line 44; Figure 4], The signal denoising module includes a reservoir computer that performs computing on input signals.)
“a reservoir main unit including a plurality of neurons connected to each other by synapses” ([col. 10, lines 53-57, Figure 4], Element 402 in Figure 4 depicts the reservoir, which contains a plurality of neurons connected together by synapses.)
“a read-out that is configured to calculate and output an inner product of a weight vector and an activity value vector, each element of which is an activity value output from each of the plurality of neurons on a basis of an input to the input layer” ([col. 10, lines 39-64; col. 11, lines 12-63; Figure 4], The reservoir computer contains a reservoir with a plurality of neurons that are connected by synapses and receives input data. Readout layers are capable of performing calculations on the input data from the reservoir and generating an output. The calculation of the ordinary differential equations by the reservoir computer shows the terms D, adaptable mixing weight vectors and u(t), input signal at time t. The function y(t) shows the inner product of the weight vector and input signal (activity value vector) in the second term of the equation. The activity value is an output from the reservoir computer after computation on the input data. The outputs 404 in Figure 4 describes these activity values.)
“a learning unit configured to acquire the observed signal from the observed signal receiver, calculate an error between the inner product and the observed signal, and update the weight vector using a value obtained by applying an adaptive filter to the error” ([col. 9, lines 59-67; col. 10, lines 1-3, lines 65-67; col. 11, lines 1-10, Figure 6], A weight adaptation module (learning unit) calculates the error between a predicted input signal (inner product) and an actual input (observed signal). The error is used by the weight adaptation module to further tune and update the weights in an iterative process. The weight adaptation module adapts the output of the reservoir via gradient descent (adaptive filtering). The Specification does not provide an explicit definition of adaptive filtering. Thus, under the broadest reasonable interpretation, adaptive filter is defined as a digital filter to adjust the weight values based on an optimization algorithm. The weight adaption module continuously tunes the weights to minimize the error between the predicted input signal and the actual input using gradient descent algorithm. From Figure 6, the actual output 618 comes from the ADC frontends (observed signal receiver).)
“a norm calculation unit configured to sequentially calculate a norm of the weight vector updated by the learning unit” ([col. 11 lines 55-67; col. 12, lines 40-53; Figure 6], From Figure 6, the output weights from the reservoir computer are fed into the weight adaption module and loop 613 shows that the weights are sequentially updated in an iterative process. The quadratic error function, E[C,D] contains the parameters of C and D, which are both output weights. The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the output weights.)
“generate a time series data of the norm” ([col. 11, lines 12-67; col. 12, lines 1-56], The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the adaptable mixing weight vectors. The output weights, C and D, are iteratively updated using a first-order approximation with a discretization step size and the short-time linear prediction of the input signal is generated for multiple time points to capture the time evolution of the new state variables. Under the broadest reasonable interpretation, the claimed invention (par. 34 in Specification) defines generating time-series data of the norm as repeatedly calculating the norm at different intervals of time. Thus, the quadratic error is determined for the short-time linear prediction at different time step of the signal and the error includes the calculation of the norms of the output weights.)
“a determination unit configured to determine whether an outlier is included in the observed signal on a basis of at least one of the norms calculated by the norm calculation unit, in a case where ” ([col. 10, lines 66-67; col. 11, lines 1-5, 55-67; col. 12, lines 40-56], The neural network shows that there is a feature anomaly (outlier) in the actual sensor signal (observed signal) when the calculated prediction error shows a huge increase. The calculation of the quadratic error function is E[C,D] and the function is computed with respect to the calculated norms of C and D, which are the output weights. When there is a spike detected in the error signal, the weight adaptation module adjusts the output weights of the dynamic reservoir to make corrections to the neural network. It is implied that a spike in the error signal indicates the error signal exceeds the average value of all previously determined error signals.)
Rao does not explicitly disclose an implementation of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more”. However, Oostendorp discloses in the same field of endeavor:
“... in a case where a moving average of the [measurements] exceeds a predetermined threshold value a predetermined number of times or more” ([0054, Figure 11], A real-time statistical process control tool generates a trend plot to monitor measurements from a part or machine. The trend plot includes one or more trend lines to show an average level, lower level, and upper level threshold. When the one or more of the plurality of points on the trend line crosses the average value more than a predetermined number of times, alert may be instantiated indicating that one or more points are fluctuating (outlier) above or below the mean value more than a predetermined number of times.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” from Oostendorp into the teaching of Rao. The statistical process control-based heuristics from Oostendorp may be implemented into Rao for monitoring the spike in the calculated prediction error for anomaly detection. Doing so can improve anomaly detection systems by monitoring data collected from one or more sensors to determine when a predetermined threshold value is exceeded. (Oostendorp, abstract).
Regarding claim 4, Rao teaches:
“An outlier detection method comprising” (abstract, Rao describes a system that receives input data and determines the noise in the data.)
“a reservoir computing step of outputting, by a read-out of a reservoir computer of an outlier detection device in a system, an inner product of a weight vector and an activity value vector, each element of which is an activity value output by each of a plurality of neurons of a reservoir main unit of the reservoir computer, the plurality of neurons being connected to each other by synapses of the reservoir main unit on a basis of an input to an input layer of the reservoir computer” ([col. 10, lines 39-64; col. 11, lines 12-63; Figure 4], The reservoir computer contains a reservoir with a plurality of neurons that are connected by synapses and receives input data. Readout layers are capable of performing calculations on the input data from the reservoir and generating an output. The calculation of the ordinary differential equations by the reservoir computer shows the terms D, adaptable mixing weight vectors and u(t), input signal at time t. The function y(t) shows the inner product of the weight vector and input signal (activity value vector) in the second term of the equation. Figure 4 shows the reservoir computer, which includes input 400, the reservoir main unit 402, and the trainable readouts providing the outputs of the computing process.)
“a learning step of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, calculating an error between the inner product and the observed signal, and updating the weight vector using a value obtained by applying an adaptive filter to the error” ([col. 9, lines 59-67; col. 10, lines 1-3, lines 65-67; col. 11, lines 1-10, Figure 6], A weight adaptation module (learning unit) calculates the error between a predicted input signal (inner product) and an actual input (observed signal). The error is used to further update the weights. The reservoir computer is also capable of applying an adaptive filter to the outputs of the trainable readouts. From Figure 6, the actual output 618 comes from the ADC frontends (observed signal receiver) and the antennae.)
“a norm calculation step of sequentially calculating a norm of the weight vector updated in the learning step” ([col. 11 lines 55-67; col. 12, lines 40-53; Figure 6], From Figure 6, the output weights from the reservoir computer are fed into the weight adaption module and loop 613 shows that the weights are sequentially updated in an iterative process. The quadratic error function, E[C,D] contains the parameters of C and D, which are both output weights. The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the output weights.)
“generating a time series data of the norm” ([col. 11, lines 12-67; col. 12, lines 1-56], The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the adaptable mixing weight vectors. The output weights, C and D, are iteratively updated using a first-order approximation with a discretization step size and the short-time linear prediction of the input signal is generated for multiple time points to capture the time evolution of the new state variables. Under the broadest reasonable interpretation, the claimed invention (par. 34 in Specification) defines generating time-series data of the norm as repeatedly calculating the norm at different intervals of time. Thus, the quadratic error is determined for the short-time linear prediction at different time step of the signal and the error includes the calculation of the norms of the output weights.)
“a determination step of determining whether an outlier is included in the observed signal on a basis of at least one of the norms calculated in the norm calculation step, in a case where ” ([col. 10, lines 66-67; col. 11, lines 1-5, 55-67; col. 12, lines 40-56], The neural network shows that there is a feature anomaly (outlier) in the actual sensor signal (observed signal) when the calculated prediction error shows a huge increase. The calculation of the quadratic error function is E[C,D] and the function is computed with respect to the calculated norms of C and D, which are the output weights. When there is a spike detected in the error signal, the weight adaptation module adjusts the output weights of the dynamic reservoir to make corrections to the neural network. It is implied that a spike in the error signal indicates the error signal exceeds the average value of all previously determined error signals.)
Rao does not explicitly disclose an implementation of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more”. However, Oostendorp discloses in the same field of endeavor:
“... in a case where a moving average of the [measurements] exceeds a predetermined threshold value a predetermined number of times or more” ([0054, Figure 11], A real-time statistical process control tool generates a trend plot to monitor measurements from a part or machine. The trend plot includes one or more trend lines to show an average level, lower level, and upper level threshold. When the one or more of the plurality of points on the trend line crosses the average value more than a predetermined number of times, alert may be instantiated indicating that one or more points are fluctuating (outlier) above or below the mean value more than a predetermined number of times.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” from Oostendorp into the teaching of Rao. The statistical process control-based heuristics from Oostendorp may be implemented into Rao for monitoring the spike in the calculated prediction error for anomaly detection. Doing so can improve anomaly detection systems by monitoring data collected from one or more sensors to determine when a predetermined threshold value is exceeded. (Oostendorp, abstract).
Regarding claim 5, Rao teaches:
“A non-transitory storage medium storing an outlier detection program causing a computer to execute” (abstract, col. 6, lines 23-32, Rao describes a system that receives input data and determines the noise in the data.)
“a reservoir computing function of having an input layer” ([col. 10, line 44; Figure 4], The signal denoising module includes a reservoir computer that performs computing on input signals.)
“a reservoir main unit including a plurality of neurons connected to each other by synapses” ([col. 10, lines 53-57, Figure 4], Element 402 in Figure 4 depicts the reservoir, which contains a plurality of neurons connected together by synapses.)
“a read-out for outputting an inner product of a weight vector and an activity value vector in an outlier detection device of a system, each element of which is an activity value output from each of the plurality of neurons on a basis of an input to the input layer” ([col. 10, lines 39-64; col. 11, lines 12-63; Figure 4], The reservoir computer contains a reservoir with a plurality of neurons that are connected by synapses and receives input data. Readout layers are capable of performing calculations on the input data from the reservoir and generating an output. The calculation of the ordinary differential equations by the reservoir computer shows the terms D, adaptable mixing weight vectors and u(t), input signal at time t. The function y(t) shows the inner product of the weight vector and input signal (activity value vector) in the second term of the equation.)
“a learning function of acquiring an observed signal from an observed signal receiver of the system that acquires the observed signal from a sensor of the system, calculating an error between the inner product and the observed signal, and updating the weight vector using a value obtained by applying an adaptive filter to the error” ([col. 9, lines 59-67; col. 10, lines 1-3, lines 65-67; col. 11, lines 1-67, col. 12, lines 1-30, Figure 6], A weight adaptation module (learning unit) calculates the error signal based on equations of the inner product and input signal. The error is used to further update the weights. The reservoir computer is also capable of applying an adaptive filter to the outputs of the trainable readouts. From Figure 6, the actual output 618 comes from the ADC frontends (observed signal receiver) and the antennae.)
“a norm calculation function of sequentially calculating a norm of the weight vector updated in the learning function” ([col. 11 lines 55-67; col. 12, lines 40-53; Figure 6], From Figure 6, the output weights from the reservoir computer are fed into the weight adaption module and loop 613 shows that the weights are sequentially updated in an iterative process. The quadratic error function, E[C,D] contains the parameters of C and D, which are both output weights. The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the output weights.)
“generating a time series data of the norm” ([col. 11, lines 12-67; col. 12, lines 1-56], The quadratic error function contains the terms,
C
(
t
)
2
and
D
(
t
)
2
, which describes a norm calculation performed on the adaptable mixing weight vectors. The output weights, C and D, are iteratively updated using a first-order approximation with a discretization step size and the short-time linear prediction of the input signal is generated for multiple time points to capture the time evolution of the new state variables. Under the broadest reasonable interpretation, the claimed invention (par. 34 in Specification) defines generating time-series data of the norm as repeatedly calculating the norm at different intervals of time. Thus, the quadratic error is determined for the short-time linear prediction at different time step of the signal and the error includes the calculation of the norms of the output weights.)
“a determination function of determining whether an outlier is included in the observed signal on a basis of at least one of the norms calculated in the norm calculation function, in a case where ” ([col. 10, lines 66-67; col. 11, lines 1-5, 55-67; col. 12, lines 40-56], The neural network shows that there is a feature anomaly (outlier) in the actual sensor signal (observed signal) when the calculated prediction error shows a huge increase. The calculation of the quadratic error function is E[C,D] and the function is computed with respect to the calculated norms of C and D, which are the output weights. When there is a spike detected in the error signal, the weight adaptation module adjusts the output weights of the dynamic reservoir to make corrections to the neural network. It is implied that a spike in the error signal indicates the error signal exceeds the average value of all previously determined error signals.)
Rao does not explicitly disclose an implementation of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more”. However, Oostendorp discloses in the same field of endeavor:
“... in a case where a moving average of the [measurements] exceeds a predetermined threshold value a predetermined number of times or more” ([0054, Figure 11], A real-time statistical process control tool generates a trend plot to monitor measurements from a part or machine. The trend plot includes one or more trend lines to show an average level, lower level, and upper level threshold. When the one or more of the plurality of points on the trend line crosses the average value more than a predetermined number of times, alert may be instantiated indicating that one or more points are fluctuating (outlier) above or below the mean value more than a predetermined number of times.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of “where a moving average of the norms exceeds a predetermined threshold value a predetermined number of times or more” from Oostendorp into the teaching of Rao. The statistical process control-based heuristics from Oostendorp may be implemented into Rao for monitoring the spike in the calculated prediction error for anomaly detection. Doing so can improve anomaly detection systems by monitoring data collected from one or more sensors to determine when a predetermined threshold value is exceeded. (Oostendorp, abstract).
Regarding claim 7, Rao teaches:
“wherein the determination unit is configured to determine that the outlier is included in the observed signal in a case where an amount of change of the norms ” ([col. 10, lines 66-67; col. 11, lines 1-5, 55-67; col. 12, lines 40-56], The calculation of the quadratic error function is E[C,D] and the function is computed with respect to the calculated norms of C and D, which are the output weights. When there is a spike detected in the error signal, the weight adaptation module adjusts the output weights of the dynamic reservoir to make corrections to the neural network. Detecting a spike in the error indicates a feature anomaly due to a sudden change in the calculated norms of the output weights.)
Rao does not explicitly disclose an implementation of “an amount of change of the norms exceeds a predetermined threshold value a predetermined number of times or more”. However, Oostendorp discloses in the same field of endeavor:
“... where an amount of change of the ” ([0054, Figure 11], A real-time statistical process control tool generates a trend plot to monitor measurements from a part or machine. The trend plot includes one or more trend lines to show an average level, lower level, and upper level threshold. When the one or more of the plurality of points on the trend line crosses the average value more than a predetermined number of times, alert may be instantiated indicating that o