DETAILED ACTION
This action is in response the communications filed on 08/27/2025 in which claims 1-6, 14-18, 20-21, 25-30, and 32 have been amended, claims 7-13, 19, 22-24, 31 and 33-95 have been canceled and therefore claims 1-6, 14-18, 20-21, 25-30 and 32 are pending.
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 .
Claim Objections
Claim 25 is objected to because of the following informalities:
In claim 25, “means for evaluating models to:” should be “means for evaluating the models to:”
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 32 is rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. Claim dependency for claim 32 is missing. Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
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.
Claim 25 recites
“means for building models to determine a sufficiency metric based on a threshold number of historical data points;
means for evaluating models to:
assign a first degree value to a first model based on the sufficiency metric;
execute the first model to determine an accuracy metric based on a first prediction corresponding to the first degree value; and
means for training to train the first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold.”
that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f). The specification identified the corresponding structure for all the “means” elements in [00194] “one processor to… determine an accuracy metric of the first model type based on a first prediction corresponding to the default features…,” [0067] “the example scheduling framework 202 of FIGS. 2A, 2B and 3A could be implemented by one or more analog or digital circuit(s), logic circuits, programmable processor(s)…,” [0068] “Flowcharts representative of example hardware logic, machine readable instructions, hardware implemented state machines, and/or any combination thereof for implementing the scheduling framework 202 of FIGS. 2A, 2B and 3A are shown in FIGS. 5A1,5A2, 5A3, 5B, 6A, 6B, 7, 8A-8E, 9 and 10. The machine readable instructions may be one or more executable programs or portion(s) of an executable program for execution by a computer processor such as the processor 812….”
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 1-6, 14-18, 20-21, 25-30 and 32 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more
Step 1: Claims 1-6 recite an apparatus comprising one processor. Claims 14-18 and 20-21 recite one non-transitory computer readable medium. Claims 25-30 and 32 recite an apparatus and properly invoke 112(f) with the corresponding structure in the specification. Therefore, claims 1-6 are directed to a machine, and claims 14-18 and 20-21 are directed to a manufacture, and claims 25-30 and 32 are directed to a machine.
With respect to claims 1, 14 and 25:
2A Prong 1: The claim recites a judicial exception.
determine a sufficiency metric based on a threshold number of historical data points (mental process – evaluation or judgement: determine a sufficiency a metric based on a threshold number of data points)
assign a first degree value to a first model based on the sufficiency metric (mental process – evaluation or judgement: assign a first degree value to a first model)
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 1) interface circuitry; machine-readable instructions; and at least one processor circuit to be programmed by the machine-readable instructions to (claim 14) machine-readable instructions that cause at least one processor circuit to at least (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
execute the first model to determine an accuracy metric based on a first prediction corresponding to the first degree value (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
train the first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 1) interface circuitry; machine-readable instructions; and at least one processor circuit to be programmed by the machine-readable instructions to (claim 14) machine-readable instructions that cause at least one processor circuit to at least (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
execute the first model to determine an accuracy metric based on a first prediction corresponding to the first degree value (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
train the first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
With respect to claims 2, 15 and 26:
2A Prong 1: The claim recites a judicial exception.
increase the accuracy metric of the first model by increasing the first degree value to the second degree value of the first model (mental process – judgement and evaluation: increase a degree feature, and as a result increase the accuracy)
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 2) one or more of the at least one processor circuit (claim 15) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 2) one or more of the at least one processor circuit (claim 15) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
With respect to claims 3 and 27:
2A Prong 2: The judicial exception is not integrated into a practical application.
wherein the first model is a polynomial regression model (Claim 1 recites the first model is executed or trained, which is mere instructions to apply. Specifying a specific regression model does not cause the limitation to integrate those steps into a practical application - MPEP 2106.05(f))
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
wherein the first model is a polynomial regression model (Claim 1 recites the first model is executed or trained, which is mere instructions to apply. Specifying a specific regression model does not cause the limitation to integrate those steps into a practical application - MPEP 2106.05(f))
With respect to claims 4, 16 and 28:
2A Prong 1: The claim recites a judicial exception.
set a polynomial activation weight to cause proportional utilization of the first model and a second model when generating predictions (mental process – evaluation or judgement: set a weight to the first model and a second model)
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 4) one or more of the at least one processor circuit (claim 16) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 4) one or more of the at least one processor circuit (claim 16) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
With respect to claims 5, 17 and 29:
2A Prong 1: The claim recites a judicial exception.
set the polynomial activation weight to a first activation value based on the sufficiency metric (mental process – evaluation or judgement: set the weight based on the sufficiency metric)
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 5) one or more of the at least one processor circuit (claim 17) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 5) one or more of the at least one processor circuit (claim 17) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
With respect to claims 6, 18 and 30:
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 6) one or more of the at least one processor circuit (claim 18) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
cause exclusive utilization of the first model and prevention of utilization of the second model based on the first activation value (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 6) one or more of the at least one processor circuit (claim 18) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
cause exclusive utilization of the first model and prevention of utilization of the second model based on the first activation value (mere instructions to apply an exception - MPEP 2106.05(f), (3) The particularity or generality of the application of the judicial exception)
With respect to claims 20 and 32:
2A Prong 1: The claim recites a judicial exception.
identify the historical data points as at least one of historical model training data or historical job-mapping data (mental process – judgement and evaluation: identifying the data as training data)
2A Prong 2: The judicial exception is not integrated into a practical application.
(claim 20) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
(claim 20) the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
With respect to claim 21:
2A Prong 1: The claim recites a judicial exception.
calculate the sufficiency metric based on prior job allocation instances to resources (mental process – evaluation: calculating the metric)
2A Prong 2: The judicial exception is not integrated into a practical application.
the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
the machine-readable instructions are to cause the at least one processor circuit (mere instructions to apply an exception - MPEP 2106.05(f), (2) invoking generic computer components)
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
Claims 1-3, 14-15, 20, 25-27 and 32 rejected under 35 U.S.C. 103 as being unpatentable over Iman ("Neural network and regression based processor load prediction for efficient scaling of Grid and Cloud resources" 20111222) in view of Zhou ("Trajectory Prediction Based on Improved Sliding Window Polynomial Fitting Prediction Method" 20161210) and in further view of Todorovski ("Inducing Polynomial Equations for Regression" 2004)
In regard to claims 1, 14 and 25, Iman teaches: An apparatus to improve job resource scheduling efficiency, comprising: interface circuitry; machine-readable instructions; and at least one processor circuit to be programmed by the machine-readable instructions to: (Iman, p. 1 "allocating resources [job resource scheduling] in advance based on prediction models could improve the quality of the service of the cloud platform. In this paper we present time delay neural network and regression methods for predicting future workload in the Grid or Cloud platform."; A grid or cloud platform is a distributed computing infrastructure that includes computers to work together. The platform inherently teaches computer components [interface circuitry, instructions, processor circuit...])
… execute the first model to determine an accuracy metric based on a first prediction corresponding to the first degree value; and (Iman, p. 3 "Local polynomial regression [the first model] has been shown to be a useful nonparametric technique in various local modelling (e.g. [18], [19]). Our regression models are built using the least square method [21]. We used single predictor for building regression models... The quantities necessary for error computation [an accuracy metric] in our regression models is given in (7)-(13). In the following, yi is the original load of ith interval, y^ is the predicted load [a first prediction] and y¯ is the mean load... Sum of squares due to error (7)... Sum of squares due to regression... (8)... Mean of Absolute percentage error MAPE... (13) "; p. 5 "To have better analysis and comparisons, we started with a degree greater than or equal to 5 [the first degree value] and continued up to degree 12.")
Iman does not teach, but Zhou teaches: determine a sufficiency metric based on a threshold number of historical data points; (Zhou, p. 204 "1) Determination of Sliding Window Width [determine a sufficiency metric] Because the prediction bases on the finite history data, we must make a choice of the time window width… If the time window width is large, and the historical data is too much... If the time window width is too small, which means the amount of data is too small to calculate out a polynomial function... Generally speaking, we select the data length that make the forecast error be minimum according to the actual measurement data. Set the historical data number n=3 [e.g. a threshold number of historical data points]"; p. 202 "Assuming a given data points {Xi|i=0,1,… ,n}")
assign a first degree value to a first model based on the sufficiency metric; (Zhou, p. 204 "2) Determination of the Number of Degree [assign a first degree value] According to the principle of least squares fitting polynomial, firstly, the degree of polynomials in the prediction model [a first model] must be determined. Determination of the number of degree should base on the analyses of the actual measurement data and the sliding window width, [the sufficiency metric] and the moving target can not be uniform due to various limitations. Set the polynomial degree m = 2.")
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Iman to incorporate the teachings of Zhou by including historical data group with the sliding window. Doing so would achieve higher prediction precision which is higher than the traditional sliding window polynomial fitting prediction method. (Zhou, p. 202 "This method improves the traditional sliding window polynomial fitting method and selects an appropriate historical data group for every predicted value, which provides every prediction with a suitable polynomial fitting equation and make the most use of the current value, and its prediction precision is higher than the traditional sliding window polynomial fitting prediction method")
Iman and Zhou do not teach, but Todorovski teaches: train the first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold. (Todorovski, p. 445 "As Ciper, the stepwise regression method also starts with the simplest model vd = const [the first degree value = 0] and sequentially adds those independent variables [the second degree value = 1, 2, ...] to the model that most significantly improve its fit to the training data. [training the first model] To avoid overfitting, stepwise regression methods test the significance of the MSE improvement [the accuracy metric] gained by refining the current equation and do not take into account those refinements that do not lead to significant improvements. The significance of the MSE improvement is based on F statistic: F = MSE (vd = P) - MSE (P') / MSE (P')... The improvement is significant, if the obtained F value is greater than the 95th percentile of the F(1, m − r − 2) distribution [9]. [continue training or searching if F > 95%, i.e. continue training or searching if F is not ≤ 95%, i.e. does not satisfy an accuracy threshold] Stepwise regression method proceed with greedy search by choosing the best significant improvement and stops, if no significant improvement is available."; p. 446 "Ciper can be viewed as a stepwise method for polynomial regression [the first model] with MDL heuristic function"; p. 451 " where v is an independent variable v ∈ V \ {vd}"; p. 442 "For example, consider a set of variables V = {x, y, z}, where z is chosen to be a dependent variable. The term x (that is equivalent to x1y0) has degree 1, the term x2y has degree 3, while x2y3 is a term of degree 5. An example polynomial equation is z = 1.2x2y + 3.5xy3. It has degree 4 and length 7.")
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Iman and Zhou to incorporate the teachings of Todorovski by including Ciper, an efficient method for inducing polynomial equations. Doing so would allow polynomials from this method to be better than linear and piecewise regression models, induced by standard regression methods, in terms of degree of fit and complexity. (Todorovski, p. 441 "We present Ciper, an efficient method for inducing polynomial equations and empirically evaluate its predictive performance on standard regression tasks. The evaluation shows that polynomials compare favorably to linear and piecewise regression models, induced by standard regression methods, in terms of degree of fit and complexity.")
Claims 14 and 25 recite substantially the same limitation as claim 1, therefore the rejection applied to claim 1 also apply to claims 14 and 25. In addition, Iman teaches: (claim 14) At least one non-transitory computer readable medium comprising machine-readable instructions that cause at least one processor circuit to at least: (Iman, p. 1 "In this paper we present time delay neural network and regression methods for predicting future workload in the Grid or Cloud platform."; A grid or cloud platform is a distributed computing infrastructure that includes computers to work together. The platform inherently teaches computer components [interface circuitry, instructions, processor circuit...])
PNG
media_image1.png
116
412
media_image1.png
Greyscale
In regard to claims 2, 15 and 26, Iman teaches: wherein one or more of the at least one processor circuit is to increase the accuracy metric of the first model by increasing the first degree value to the second degree value of the first model. (Iman, p. 5 "B. Polynomial Regression Model [the first model]... To have better analysis and comparisons, we started with a degree greater than or equal to 5 and continued up to degree 12. The best result is achieved by the degree 8 model followed by degree 6 and degree 5 model."; p. 5 "Table VII Polynomial degree 5... 29.66... Polynomial degree 8... 23.12"; the more the degree value (5 -> 8) the less the error (accuracy increased))
In regard to claims 3 and 27, Iman teaches: wherein the first model is a polynomial regression model. (Iman, p. 1 "In this paper we present time delay neural network and regression methods for predicting future workload in the Grid or Cloud platform."; p. 5 "B. Polynomial Regression Model…")
In regard to claims 20 and 32, Iman does not teach, but Zhou teaches: wherein the machine-readable instructions are to cause the at least one processor circuit to identify the historical data points as at least one of historical model training data or historical job-mapping data. (Zhou, p. 202 "This method improves the traditional sliding window polynomial fitting method and selects an appropriate historical data group for every predicted value, which provides every prediction with a suitable polynomial fitting equation and make the most use of the current value..."; p. 204 "When the improved sliding window polynomial fitting prediction method predicts the multiple consecutive future value simultaneously, an appropriate historical data group will be selected aiming at every predicted value, the specific method is to regard the distance between the future value and the current value as sampling interval, and acquired the historical data group, including the current value, by sampling the historical data set according to the sliding window width."; p. 204 "Set the historical data number n=3, the prediction number q=3."; historical data are used for polynomial fitting prediction, i.e. historical data are used as training data)
The rationale for combining the teachings of Iman and Zhou is the same as set forth in the rejection of claim 1.
Claims 4, 16 and 28 rejected under 35 U.S.C. 103 as being unpatentable over Iman, Zhou and Todorovski as applied to claims 1, 14 and 25, and in further view of Bonab ("Less Is More: A Comprehensive Framework for the Number of Components of Ensemble Classifiers" 20190109)
In regard to claims 4, 16 and 28, Iman, Zhou and Todorovski do not teach, but Bonab teaches: wherein one or more of the at least one processor circuit is to set a polynomial activation weight to cause proportional utilization of the first model and a second model when generating predictions. (Bonab, p. 2738, B. Weighted Majority Voting (WMV) "For this aggregation rule, a weight vector w=⟨W1,W2,…,Wm⟩ for components of ensemble is defined, Wj≥0 and ∑Wj=1 for 1≤j≤m..."; e.g. W1 = 0.7, W2 = 0.3, W1+W2 = 1)
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Iman, Zhou and Todorovski to incorporate the teachings of Bonab by including a flexible aggregation rule. Doing so would allow cancelling a poor component’s effect on the aggregated results. (Bonab, p. 2738, "Note that giving equal weights to all the components will result in the MV aggregation rule. WMV presents a flexible aggregation rule. No matter how poor a component classifier is, with a proper weight vector we can cancel its effect on the aggregated results.")
Claims 5-6, 17-18 and 29-30 rejected under 35 U.S.C. 103 as being unpatentable over Iman, Zhou and Todorovski as applied to claims 4, 16 and 28, and in further view of Rahmanian ("A learning automata-based ensemble resource usage prediction algorithm for cloud computing environment" 20171012)
In regard to claims 5, 17 and 29, Iman, Zhou, Todorovski and Bonab do not teach, but Rahmanian teaches: wherein one or more of the at least one processor circuit is to set the polynomial activation weight to a first activation value based on the sufficiency metric. (Rahmanian, p. 59 "Algorithm 1 describes the proposed ensemble resource usage prediction algorithm by LA theory for cloud computing environment. The actions of the LA are ‘increase weight’, ‘decrease weight’, and ‘no-change weight’ actions... Over a period, the resource usage of the ith cloud resource according to its resource usage history by the jth constituent prediction model at time slot Δt [based on the sufficiency metric] for the next time slot is predicted (Line 9–10). After that, the feedback of the selected action on the environment by evaluating the differences between the actual resource usage (yi) and the predicted resource usage (y′ i,j) is evaluated (Line 11). According to the feedback of LA, the action probabilities (Pi,j) are updated (Line 12) and the optimal action in the time slot Δt is selected based on the updated probabilities (see line 13). Finally, the weight of the jth constituent prediction model for the ith cloud resource (Ai,j) is updated [set the polynomial activation weight to a first activation value] and the ensemble prediction algorithm for the ith cloud resource for the next time slot is determined (Lines 15–16).")
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Iman, Zhou and Todorovski to incorporate the teachings of Rahmanian by including Learning Automata (LA) theory to determine weight for each model. Doing so would achieve better performance than other ensemble prediction algorithms. (Rahmanian, p. 54 "In this paper, an ensemble cloud resource usage prediction algorithm based on Learning Automata (LA) theory is proposed that combines state of the art prediction models, and it determines weights for individual constituent models. The proposed algorithm predicts by combining the prediction values of all constituent models based on their performance. The extensive experiments on CPU load prediction of several VMs gathered from the dataset of the CoMon project indicated that the proposed approach outperforms other ensemble prediction algorithms.")
In regard to claims 6, 18 and 30, Iman, Zhou and Todorovski do not teach, but Bonab teaches: wherein one or more of the at least one processor circuit is to cause exclusive utilization of the first model and prevention of utilization of the second model based on the first activation value. (Bonab, p. 2738, B. Weighted Majority Voting (WMV) "For this aggregation rule, a weight vector w=⟨W1,W2,…,Wm⟩ for components of ensemble is defined, Wj≥0 and ∑Wj=1 for 1≤j≤m... Assume that the least loss belongs to component j, among m score-vectors... Simply giving a weight of 1 to j’s component [the first activation value causes exclusive utilization of the first model type] and 0 for the remaining components result in the equality case..."; in light of [0090] "the model evaluator 212 sets a polynomial activation weight value to one (e.g., 1.0) to indicate that predictions should occur exclusively by polynomial regression modeling approaches, and prevents utilization of any other model type (e.g., LSTM)")
The rationale for combining the teachings of Iman, Zhou, Todorovski and Bonab is the same as set forth in the rejection of claim 4.
Claim 21 rejected under 35 U.S.C. 103 as being unpatentable over Iman, Zhou and Todorovski as applied to claim 14, and in further view of Baig ("Adaptive Prediction Models for Data Center Resources Utilization Estimation" 20190802)
PNG
media_image2.png
276
373
media_image2.png
Greyscale
In regard to claim 21, Iman, Zhou and Todorovski do not teach, but Baig teaches: wherein the machine-readable instructions are to cause the at least one processor circuit to calculate the sufficiency metric based on prior job allocation instances to resources. (Baig, p. 1691 "Figure 14 shows the RMSE and MAE for different windows sizes with the proposed system to estimate CPU resource estimation. We observe that increasing window size reduce the estimation error until window size 60; however, after that, the error starts rising. The 20 minutes window size only contains four observations to fit the prediction models for an estimation which yields a maximum error. This experiment identifies that 60 minutes window size is optimal [calculate the sufficiency metric (window size)] to use with the proposed system to minimize the estimation error. Therefore, in all of our experiments, we used 60 minutes window size with the proposed and baseline methods."; p. 1681 "The proposed approach trains a classifier based on statistical features of historical resources usage [prior job allocation instances to resources] to decide the appropriate prediction model"; p. 1687 "The Google cluster traces [48] are the publicly available traces published by Google. To create the CPU and the Memory utilization, the tasks of each job were aggregated by summing their CPU and Memory consumption every five minutes in a period of 24 hours.")
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to have modified Iman, Zhou and Todorovski to incorporate the teachings of Baig by including the RMSE and MAE for different windows sizes. Doing so would allow identifying the optimal window size to use with the proposed system to minimize the estimation error. (Baig, p. 1691 "Figure 14 shows the RMSE and MAE for different windows sizes with the proposed system to estimate CPU resource estimation. We observe that increasing window size reduce the estimation error untill window size 60; however, after that, the error starts rising. The 20 minutes window size only contains four observations to fit the prediction models for an estimation which yields a maximum error. This experiment identifies that 60 minutes window size is optimal to use with the proposed system to minimize the estimation error.")
Response to Arguments
Applicant's arguments with respect to the rejection of the claims under 35 U.S.C. 101 have been fully considered but they are not persuasive:
Applicant argues: (p. 9) Prong1… Similar to USPTO Example 39, independent claim 1 sets forth at least one processor circuit to train a first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold. The USPTO Memorandum clarifies that "Even though "training the neural network" involves a broad array of techniques and/or activities that may involve or rely upon mathematical concepts, the limitation does not set forth or describe any mathematical relationships, calculations, formulas, or equations using words or mathematical symbols."
Examiner answers: Example 39 does not recite any abstract idea at all, however, the claim in this application recites “determine a sufficiency metric based on a threshold number” and “assign a first degree value” which are abstract ideas. Further example 39 doesn’t state about what to do when a claim recites both 1) abstract idea and 2) additional elements.
The limitation “train the first model based on a second degree value when the accuracy metric does not satisfy an accuracy threshold” is not an abstract idea, instead it is evaluated as an additional element which is mere instructions to apply an exception - MPEP 2106.05(f).
Applicant argues: (p. 10) Prong 2… The originally-filed application discloses, in part, technical considerations to… [0024]… [0025]… [0027]… Because the claimed subject matter "reflects an improvement to the functioning of a computer or to another technology or technical field" (USPTO Memorandum, p. 4), claim 1 is precisely on-point with the type of statutory subject matter highlighted by the recent USPTO reminder on evaluating subject matter eligibility.
Examiner answers: The amended claim 1 recites determining a sufficiency metric, and executing and training the first model, which are not in line with the features described in the [0024] [0025] [0027] (which describes using the first/polynomial model when historical data is not available, and using LSTM model when historical model is available) The features in amended claim 1 (executing and training the first model) are mere instructions to apply which doesn’t reflect improvements in the claim.
Applicant's arguments with respect to the rejection of the claims under 35 U.S.C. 103 have been fully considered but they are moot:
Applicant argues: (p. 12) because Hauser is devoid of any consideration of a threshold number of historical data points, Hauser necessarily fails to teach or suggest at least one processor circuit to determine a sufficiency metric based on a threshold number of historical data points, as set forth in claim 1… this mere indication is neither an assignment nor sufficient to teach or suggest at least one processor circuit to assign a first degree value to a first model based on the sufficiency metric, as set forth in claim 1.
Examiner answers: the arguments do not apply to the references (Zhou) being used in the current rejection.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Barak ("Smoothing and Differentiation by an Adaptive-Degree Polynomial Filter" 1995) teaches (Barak, p. 2758 "Extension of the least-squares regression formalism with statistical testing of additional terms of polynomial degree to a heuristically chosen minimum for each data window leads to an adaptive-degree polynomial filter (ADPF).")
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 SU-TING CHUANG whose telephone number is (408)918-7519. The examiner can normally be reached Monday - Thursday 8-5 PT.
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, Andrew J. Jung can be reached on (571) 270-3779. 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.C./Examiner, Art Unit 2146
/ANDREW J JUNG/Supervisory Patent Examiner, Art Unit 2146