DETAILED ACTION
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 .
Information Disclosure Statement
The information disclosure statement (IDS) was submitted on 10/31/2023. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Specification
The disclosure is objected to because of the following informalities:
Specification paragraph [0001] reads: “scattered light provides a spectrum indicated of the sample components.” Examiner suggests the text should read “scattered light provides a spectrum indicative of the sample components.”
Appropriate correction is required.
Claim Objections
Claim 19 objected to because of the following informalities:
Claim 19 reads: “… scientific instrument support apparatus according to claim 18, the computing device is implemented in a scientific instrument…”. Examiner suggest the claim is missing the word “wherein” and should read: “…scientific instrument support apparatus according to claim 18, wherein the computing device is implemented in a scientific instrument…”, to be consistent with language of other claims.
Claim 28 objected to because of the following informalities:
Claim 28 reads: “to provide a trained models;”. Examiner suggests the word “a” is misplaced as written and claim should read “to provide trained models;” or alternatively, “to provide a set of trained models;”, depending on Applicant’s intended meaning. For examination purposes, the limitation will be interpreted to mean the method provides a plurality of trained models.
Appropriate correction is required.
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 1-3, 9, 12-15, 17-19, and 21-29are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. These claims fall into statutory categories as set forth in 35 U.S.C. 101 (See MPEP § 2106.03), as discussed in detail below.
Specifically, independent Claim 1 recites:
“scientific instrument support apparatus, comprising: a first logic to manage and pre-process a spectroscopic data set; a second logic to train one or more models and provide one or more of a trained model; and a third logic to provide a measure of a quality of the one or more trained models and provide a one or more of a found hyperparameter of the one or more trained models.”
(Examiner notes emphasis added above for clarity in discussion)
Claim 1 is considered to be in a statutory category: machine or manufacture, based on Step one of eligibility analysis. (MPEP § 2106.03). Next, evaluation under Step 2A, Prong One, and applying broadest reasonable interpretation, reveals the limitation recites a judicial exception of abstract idea in the “Mathematical Concept” grouping. (See MPEP §2106.04(a)(2), subsection I.)
Specifically, Examiner points to text emphasized in bold in Claim 1 as
shown above. The terms “first logic”, “second logic”, and “third logic” are interpreted to mean nothing more than mathematical or computational process steps using input values and providing output values. Interpretation of these terms to mean mathematical or computational steps is confirmed in consideration of terms emphasized in italics which follow the mathematical steps, each of which describe either a mathematical procedure or operation as with terms including “manage and pre-process”, and “train”, or a describe an output or result of a procedure or operation, as with term “provide”.
Claim limitation language with these terms is interpreted as nothing more than a series of mathematical calculations using input data which results in output data or values, as shown with emphasis in italic underline, as “spectroscopic data set” and “models” (input values), and which results in output models, data, or values, which include: one or more of a trained model, a measure of a quality, trained models and hyperparameter.
This interpretation and conclusion is supported by further review of the specification, for example, specification beginning in [0004], where the mathematical processes are discussed, and in [0130] where mathematical concepts are described as “logic of the scientific instrument support module 1000 may be included in a single computing device or may be distributed across multiple computing devices”; with additional guidance from [0131] , reciting “used herein, the term "logic" may include an apparatus that is to perform a set of operations associated with the logic…any of the logic elements included in the support module 1000 may be implemented by one or more computing devices programmed with instructions to cause one or more processing devices of the computing devices to perform the associated set of operations”
In further evaluation of eligibility, Step 2A, Prong two was applied, with
consideration as to whether the limitations of Claim 1 recite additional elements
that may integrate the judicial exception into practical application. (MPEP §
2106.04(d)(2)) Specifically, Examiner notes the following additional element in Claim 1 emphasized above in underline as “scientific instrument support apparatus”. Examiner considers the apparatus as an additional element referring to an instrument for gathering of numerical values to be used in claimed mathematical or computational operations and/or computational support instrumentation for carrying out claimed mathematical or computational operations, and as such is considered to be insignificant extra solution activity. The terms are interpreted as generally known or mere data gathering, necessary to provide numerical values based on measuring, as would be known by one of ordinary skill in the art, for use in recited mathematical operations related to the calculations recited above. This interpretation is supported in further review of the specification, as noted above. The acquisition of data by measuring, intended to be used in the cited mathematical concept is not meaningful because this represents nothing more than a necessary precursor required to carry out mathematical calculation. (see MPEP 2106.05(g), 2106.05(f)) Further, location and arrangement of acquiring measurements necessary as input data for mathematical calculations, for example the location of monitoring stations, simply recite field of use/technological environment (see MPEP 2106.05(h))
Thus, Claim 1 does not include additional elements that are sufficient to
amount to significantly more than the judicial exception because these additional
elements/steps are well known, routine and conventional as evidenced by in the relevant art based on the prior art of record cited herein, including, for example: (Patent literature) BRULLOT (US 20220003679 A1), BAUER (US 20220223230 A1, LUNDSTEDT (US 7194369 B2) and (Non-Patent Literature) SUI (SUI, et al., "A deep learning model designed for Raman spectroscopy with a novel hyperparameter optimization method", Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 280 (2022) 121560 - Available online 25 June 2022.) Accordingly, additional elements do not integrate the abstract idea into a practical application because the elements to not impose meaningful limits on practicing the abstract idea.
Further, the examiner does not view this claim as improving the functioning
of a computer, or improvement to any other technology or technical field. (see MPEP 2106.05(b)), nor effecting a transformation or reduction of a particular article to a different state or thing. (see MPEP 2106.05(c)). The limit does not apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception. (see MPEP 2106.05(e) and Vanda Memo).
Similarly, independent Claim 26 recites:
“Raman spectrometer comprising:
a support apparatus including; first logic to manage and pre-process spectroscopic data sets, second logic to train one or more models and provide a one or more trained model, a third logic to provide a measure of the quality of the one or more trained model and provide a one or more found hyperparameter of the one or more trained model, and a fifth logic to manage an application of the trained models to a test sample data to identify, a qualitative or quantitative feature of one or more substances in the test sample.”
Claim 26 is considered to be in a statutory category: machine or manufacture using Step one of eligibility analysis. (MPEP § 2106.03). Similarly to Claim 1, Evaluation under Step 2A, Prong One, and applying broadest reasonable interpretation, reveals the limitation recites a judicial exception of abstract idea in the “Mathematical Concept” grouping. (See MPEP §2106.04(a)(2), subsection I.)
Using the same reasoning as applied to terms discussed above for Claim 1, Examiner points to text emphasized in bold in Claim 26: “first logic”, “second logic”, “third logic” and “fifth logic” which are likewise directed to specific mathematical or computational process steps using input values and providing output values. Following reasoning used in Claim 1, interpretation of these terms as mathematical or computational steps is confirmed in consideration of terms emphasized in italics which follow the mathematical steps, including terms: “train”, “provide”, “manage…test”, and “identify”, each of which describe either a mathematical procedure or operation as discussed above. Claim limitation language with these terms is interpreted as nothing more than a series of mathematical calculations using input data or output data or values, noted by terms emphasized with italic underline, including “spectroscopic data sets”, “models”, and “test sample” (input), and terms “trained model”, “a measure of the quality”, “hyperparameter”, “qualitative or quantitative feature” (output).This interpretation and conclusion is supported by further review of the specification, as noted above in Claim 1 discussion.
In further evaluation of eligibility, Step 2A, Prong two was applied, with consideration as to whether the limitations of Claim 26 recite additional elements that may integrate the judicial exception into practical application. (MPEP § 2106.04(d)(2)) Specifically, Examiner notes the following additional elements in Claim 26: “Raman spectrometer” and a “support apparatus”. Examiner considers these additional elements, as discussed above as referring to instrument or computational interface for data gathering, and is interpreted to be insignificant extra solution activity. The terms are interpreted as generally known or mere data gathering, necessary to provide numerical values based on measuring, as would be known by one of ordinary skill in the art, for use in recited mathematical operations related to calculations recited above. This interpretation is supported in further review of the specification, as noted above. The acquisition of data by measuring, intended to be used in the cited mathematical concept is not meaningful because this represents nothing more than a necessary precursor required to carry out mathematical calculation. (see MPEP 2106.05(g), 2106.05(f)) Further, location and arrangement of acquiring measurements necessary as input data for mathematical calculations, for example the location of monitoring stations, simply recite field of use/technological environment (see MPEP 2106.05(h))
Thus, Claim 26 does not include additional elements that are sufficient to
amount to significantly more than the judicial exception because these additional elements/steps are well known, routine and conventional as evidenced by in the relevant art based on the prior art of record cited herein, and cited above.
Accordingly, additional elements do not integrate the abstract idea into a practical application because the elements to not impose meaningful limits on practicing the abstract idea. Further, the examiner does not view this claim as improving the functioning of a computer, or improvement to any other technology or technical field. (see MPEP 2106.05(b)), nor effecting a transformation or reduction of a particular article
to a different state or thing. (see MPEP 2106.05(c)). The limit does not apply or use
the judicial exception in some other meaningful way beyond generally linking the
use of the judicial exception to a particular technological environment, such that the
claim as a whole is more than a drafting effort designed to monopolize the exception. (see MPEP 2106.05(e) and Vanda Memo).
Similarly, independent Claim 28 recites:
“method for scientific instrument support, comprising: managing and pre-processing data; training one or more models to provide a trained models; providing a measure of the quality of the trained model; and providing a one or more hyperparameter of the trained model.”
Claim 28 is considered to be in a statutory category: process (method), using Step one of eligibility analysis. (MPEP § 2106.03). Similarly to Claim 1, Evaluation under Step 2A, Prong One, and applying broadest reasonable interpretation, reveals the limitation recites a judicial exception of abstract idea in the “Mathematical Concept” grouping. (See MPEP §2106.04(a)(2), subsection I.)
Using the same reasoning as applied to terms discussed above for Claim 1, Examiner points to text emphasized in bold in Claim 28: “managing” and “pre-processing” which are likewise directed to specific mathematical or computational process steps using input values and providing output values. Following reasoning used in Claim 1, interpretation of these terms as mathematical or computational steps is confirmed in consideration of terms emphasized in italics which follow the mathematical steps, including terms: “train”, “provide”, and “providing”, each of which describe either a mathematical procedure or operation as discussed above. Claim limitation language with these terms is interpreted as nothing more than a series of mathematical calculations using input data or output data or values, noted by terms emphasized with italic underline, including “data”, and “models” (input), and terms “trained models”, “measure of the quality”, “hyperparameter” (output).This interpretation and conclusion is supported by further review of the specification, as noted above in Claim 1 discussion.
Using the same reasoning as above for Claims X and XX, Examiner points to text emphasized in bold in Claim XXX as shown above, citing terms which again are directed to a specific mathematical calculations involving direct or comparative analysis, resulting in either a “true/false” result or a numerical value. ETC
In consideration of dependent Claims 2-3, 9, 12-15, 17-19, 21-25, and 27, with dependency to Claim 1, and Claim 29 with dependency to Claim 28, Examiner carried out further evaluation to determine whether dependent claims recited additional elements. Examiner find dependent claim limitations recite additional extra-solution, data-gathering (Claims 2, 3, 13, 14, 19, 23, 25, ) or additional mathematical calculations (Claim 9, 12, 14, 15, 17, 18, 21, 22, 23, 24, 25, 28, 29). However, language recited in these dependent claims is not sufficient to amount to significantly more than the judicial exception. The additional elements represent insignificant extra-solution activity, or provide additional features/steps which are part of an expanded abstract idea as recited in independent Claims 1 and 9.
Examiner further notes Claim 27 recited specifically use of a Raman spectrometer comprising a support apparatus as described in previous limitations. This additional element does no more than generally link the use of the abstract idea to a particular technological environment or field of use, with specification of where input values are acquired, but does not modify the described mathematical concepts/calculations/processes in such a way that integrates the abstract idea into a practical application.
When analyzed independently or in combination, dependent Claims 2-3, 9, 12-15, 17-19, 21-25, and 27, with dependency to Claim 1, and Claim 29 with dependency to Claim 28, are held to be patent ineligible under 35 U.S.C. 101 because the additional recited limitations therein, as described above, represent additional elements that describe insignificant extra solution activity, additional mathematical calculations, instructions or definitions for calculations, and/or numerical or data to be used as input or output according to the recitation of the abstract ideas as discussed above for independent Claims 1, 26, and 28.
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.
Examiner notes application of guidance found in MPEP 2141 in determination of obviousness under 35 U.S.C. 103. Specifically, factual inquiry steps described in 2141 (II): “An invention that would have been obvious to a person of ordinary skill at the relevant time is not patentable. See 35 U.S.C. 103 or pre-AIA 35 U.S.C. 103(a). As reiterated by the Supreme Court in KSR, the framework for the objective analysis for determining obviousness under 35 U.S.C. 103 is stated in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966). Obviousness is a question of law based on underlying factual inquiries.” Further, the following steps for factual inquiries were used in evaluation of prior art used for obviousness rejection:
(A) Determining the scope and content of the prior art;
(B) Ascertaining the differences between the claimed invention and the prior art; and
(C) Resolving the level of ordinary skill in the pertinent art.
Claims 1-2, 12, 14-15, 18-19, 21-22, 27-29 are rejected under 35 U.S.C. 103(a) as being unpatentable over BRULLOT (US 20220003679 A1) in view of SUI (SUI, et al., "A deep learning model designed for Raman spectroscopy with a novel hyperparameter optimization method", Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 280 (2022) 121560 - Available online 25 June 2022).
With respect to Claim 1, BRULLOT teaches:
A scientific instrument support apparatus,
(For context, see [0001] “The present disclosure relates to a method of predicting physical properties, in particular performance parameters”, BRULLOT is in same technical area; BRULLOT teaches an computer and processor connected to a Raman spectrometer forming an automated instrument for experimental data collection, see [0115]-[0117]: “disclosure further provides a computer system for predicting at least one physical property…computer system comprising at least: [0116] an interface component configured to access and read a Raman spectrum…[0117] a processor unit implementing a model and configured to use the Raman spectrum provided via the interface component as input for the model which determines a value of the physical property from the spectroscopic data”, and see [0119]: “system is configured to be coupled to a Raman spectrometer via a wired and/or wireless communication connection, and to access and read out the Raman spectrum at least partly automatically from the Raman spectrometer via the interface component”; Examiner notes application of broadest reasonable interpretation (“BRI”) for claim limitation language “apparatus” as analogous to reference of a “system” for data acquisition in a scientific setting.)
comprising:
a first logic to manage and pre-process a spectroscopic data set;
(BRULLOT teaches a pre-processing step using spectroscopic data from a Raman spectrometer, see [0045]: “spectroscopic data have been pre-treated by baseline correction of each Raman spectrum, optional smoothing or generation of a derivative of each Raman spectrum, and subsequent normalization of the Raman spectra”, and see [0046]: “spectra are pre-treated on the x-axis and on the y-axis… Combinations of these pre-treatments are optimized depending on the product that is being characterized.”, and [0047]: “baseline correction”; Examiner notes application of BRI to claim limitation language of “first logic” to mean generally a computational process step for data handling.”)
a second logic to train one or more models and provide one or more of a trained model;
(BRULLOT teaches training models, see [0043]: “To ensure robust performance of the predictive model, continuous validation is used in one embodiment of the method. This is done in silico by determining the fit of any newly measured spectrum to the spectra used to train the model and also in the laboratory by comparing the predicted results with actual measured data”; BRULLOT teaches development of trained models for data analysis, see [0037]: “model has been trained with Raman spectra and measured values of the at least one physical property, in particular performance parameter” and see [0051]: “several models are made/trained and tested/validated with process samples”; and see Claim 2; Examiner notes again BRI as applied to claim limitation language of “second logic” to mean generally a computational process step for data handling.)
and a third logic to provide a measure of a quality of the one or more trained models
(BRULLOT teaches analysis of trained models, see [0051] During the modelling/training step of the PLS model, various model performance parameters or model KPIs are optimized: Root Mean Square Error of Cross Validation (RMSECV), Root Mean Square Error of Prediction (RMSEP), R.sup.2, Spectral Residuals, and Mahalanobis distance. Typically, several models are made/trained and tested/validated with process samples. The best model is chosen based on the reliability, accuracy and precision of the predicted/calculated values of the product performance parameters.; Examiner notes BRULLOT does teach use of “hyperparameter”, interpreted herein as guided by specification FIG. 25 and [0015] to mean, in the context of training models and data analysis, to mean generally any external configuration variable setting, the value of which is set prior to model training, and which is used to manage model training or control learning process, as would be understood by one of ordinary skill in the art, but does not explicitly teach providing hyperparameter to training model.)
However, BRULLOT is silent to the language of
[a third logic to] provide a one or more of a found hyperparameter of the one or more trained models.
Nevertheless, SUI teaches
[a third logic to] provide a one or more of a found hyperparameter of the one or more trained models.
(For context, see Pg2§1. Introduction:” deep learning model designed for one dimensional Raman spectral data and a hyperparameter optimization method for deep learning modeling”, thus, SUI is in same technical field.; SUI teaches use of hyperparameter optimization to train data analysis models, see FIG. 1 B) for data analysis steps, and see Pg.2: “hyperparameter optimization was implemented by cross-validation technique and Bayesian optimization, which made the results more accurate.” and teaches hyperparameter selection, “a designed searcher will search for the optimal combination of hyperparameters so as to obtain the most suitable model for a specific task.”, and see Fig. 2 for process steps in using hyperparameter to train models.)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to modify BRULLOT to include providing a one or more of a found hyperparameter of the one or more trained models, such as that of SUI.
One of ordinary skill would be motivated to modify BRULLOT to include providing a one or more of a found hyperparameter of the one or more trained models, as taught by SUI because it would be understood as a way to logically implement the use of hyperparameter optimization that was suggested in BRULLOT. One of ordinary skill would understand the known technique and value of using hyperparameter selection to train models for data analysis as a way to make analysis of large data sets more efficient and more accurate minimizing computational time and effort to arrive at a best model for given data. One of ordinary skill would be further motivated to use the machine learning/model training method as suggested by BRULLOT in combination with the detailed analysis techniques taught by SUI (and others, as included below in conclusion) because such processes are known improve the ability to achieve faster, more accurate and reliable quantitative and/or qualitative and/or identification information of a wide range of materials when used in the context of Raman spectroscopic data. One of ordinary skill would be motivated to use hyperparameters, as suggested by BRULLOT, and refined in the method of SUI with a reasonable expectation of success in ensuring models perform optimally and provide maximum predictive power, which would be advantageous particularly in an industrial setting.
With respect to Claim 2, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
BRULLOT further teaches:
wherein the spectroscopic data set includes Raman data from measurements of different training samples.
(BRULLOT teaches using Raman data for training models, see as above, [0012]: “In the chemometric model of the present disclosure, spectroscopic data (typically Raman scattering spectra) are correlated with physical properties, in particular performance parameters”, and see [0037]: “model has been trained with Raman spectra and measured values of the at least one physical property, in particular performance parameter”)
With respect to Claim 12, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
BRULLOT further teaches:
wherein the first logic preprocesses the spectroscopic data by applying a wavelength normalization.
(BRULLOT teaches wavelength normalization as part of pre-treatment, refer to FIG. 1 with [0013]: FIG. 1 shows an example of a set of 10 Raman spectra from a superabsorbent polymer with a linear fit subtraction as baseline correction and a SNV (standard normal variate) normalization step as data pre-treatment”; Examiner notes one of ordinary skill would know that SNV normalization is known as a wavelength normalization technique.)
With respect to Claim 14, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
BRULLOT further teaches:
wherein the model is input as a selection from different model types by the second logic.
(BRULLOT teaches model selection from options, see [0042]: “calibration models obtained in the first step are validated using independent test samples (batches that were not used in the training phase). The calibration model yielding the best results in the validation in terms of error on prediction, reduced complexity, robustness etc. is then chosen to predict the performance parameters of other samples.”; and see [0051]: “ During the modelling/training step of the PLS model, various model performance parameters or model KPIs are optimized: Root Mean Square Error of Cross Validation (RMSECV), Root Mean Square Error of Prediction (RMSEP), R2, Spectral Residuals, and Mahalanobis distance. Typically, several models are made/trained and tested/validated with process samples. The best model is chosen based on the reliability, accuracy and precision of the predicted/calculated values of the product performance parameters.)
With respect to Claim 15, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
However, BRULLOT, as modified by SUI and as taught above, is silent to the language of:
wherein the second logic trains the one or more models by Bayesian Optimization to determine the hyperparameters.
Nevertheless, SUI further teaches:
wherein the second logic trains the one or more models by Bayesian Optimization to determine the hyperparameters.
(SUI teaches use of Bayesian optimization, see Page 2, “hyperparameter optimization was implemented by cross-validation technique and Bayesian optimization, which made the results more accurate.”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify BRULLOT to include wherein the second logic trains the one or more models by Bayesian Optimization to determine the hyperparameters, such as that further disclosed by SUI.
One of ordinary skill would be motivated to further modify BRULLOT to include wherein the second logic trains the one or more models by Bayesian Optimization to determine the hyperparameters, as further taught by SUI because it would be understood improve the hyperparameter optimization as suggested in BRULLOT. One of ordinary skill would understand the advantage of using well-known Bayesian methods to determine hyperparameters in allowing for better model performance, and reducing training runs through judicious selection of a minimum number of hyperparameters. One of ordinary skill would understand and appreciate the combination of using Bayesian optimization as taught by SUI with the method disclosed by BRULLOT would result in improved efficiency for training and improved accuracy of models.
With respect to Claim 18, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
BRULLOT further teaches:
wherein the first logic, the second logic, and the third logic are implemented by a computing device.
(BRULLOT teaches computer based system and method, as above, [0012] A computer implemented method for predicting physical properties”)
With respect to Claim 19, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 18,
(See above, references as applied to Claim 18.)
BRULLOT further teaches:
the computing device is implemented in a scientific instrument, wherein the scientific instrument can measure sample spectroscopic data.
(See as above for parallel limitation in Claim 1, BRULLOT teaches computer-based data acquisition system and method, see [0115]-[0117] and [0119]: “system is configured to be coupled to a Raman spectrometer via a wired and/or wireless communication connection, and to access and read out the Raman spectrum at least partly automatically from the Raman spectrometer via the interface component”;)
With respect to Claim 21, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
However, BRULLOT, as modified by SUI and as taught above, is silent to the language of:
further comprising a fourth logic, wherein the fourth logic accepts the found
hyperparameters and trains the one or more models.
Nevertheless, SUI further teaches:
further comprising a fourth logic, wherein the fourth logic accepts the found
hyperparameters and trains the one or more models.
(As above, SUI teaches formal theoretical details for use of hyperparameter optimization, see FIG. 1 B) ; specifically SUI teaches using optimized hyperparameters for model training, see Pg. 2 Col. 2: “a designed searcher will search for the optimal combination of hyperparameters so as to obtain the most suitable model for a specific task.”; and see Pg.6Col2 §3.1: “hyperparameter optimization method is then used to learn the intrinsic relationship between Raman spectra and their pH and chip labels on the training set and evaluation is implemented on the testing set.”; and see Pg.7 Col2 §4: “we demonstrate a novel deep learning model with a hyperparameter optimization method, aiming to build the model more effectively.”)
It would have been obvious to one of ordinary skill in the art before effective filing
date of the claimed invention to further modify BRULLOT to include further comprising a fourth logic, wherein the fourth logic accepts the found hyperparameters and trains the one or more models, such as that further disclosed by SUI.
One of ordinary skill would be motivated to further modify BRULLOT to include further comprising a fourth logic, wherein the fourth logic accepts the found hyperparameters and trains the one or more models, as further taught by SUI because it would be understood as a logical next step in the use of optimized hyperparameters in the context of model training for data analysis. One of ordinary skill would understand that using identified (optimized) hyperparameters to train models would improve the overall efficiency and accuracy of a model training process.
With respect to Claim 22, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 21,
(See above, references as applied to Claim 21.)
BRULLOT further teaches:
wherein the first logic, the second logic, and the third logic are implemented on a first computing device, and the fourth logic is implemented on a second computing device.
(BRULLOT teaches implementation of system on several computing instruments, see [0115]: “disclosure further provides a computer system for predicting at least one physical property of a superabsorbent polymer, the computer system comprising at least: [0116] an interface component configured to access and read a Raman spectrum of the superabsorbent polymer; [0117] a processor unit implementing a model and configured to use the Raman spectrum provided via the interface component as input for the model which determines a value of the physical property from the spectroscopic data”; Examiner asserts BRULLOT teaches implementation of at least two computers to carry out disclosed method/system.)
With respect to Claim 27, BRULLOT, in view of SUI, teaches:
A method to identify, authenticate or quantify one or more substances in a sample under test, the method comprising:
(See as above parallel limitation in Claim 1, BRULLOT teaches system and method for using Raman spectroscopy to characterize samples in an iterative fashion using trained models for data analysis, see for example [0122]: “system may be further configured to update the model over and over again by iteratively training the model with newly measured Raman spectra and newly measured values of the at least one physical property of superabsorbent polymers.” )
irradiating the sample with an excitation beam from a spectroscopy device;
(Examiner asserts one of ordinary skill would understand that data acquisition using the Raman spectroscopy technique requires use of a well-characterized excitation source, most generally, a laser. BRULLOT teaches such a method, see [0125]: “Raman Method used to Characterize SAP”, and [0126] “Raman spectroscopy was used to characterize superabsorbent polymers…in-line application was also set up with a RamanRXN2™…laser and detector were connected to an insertion probe with a PhAT probe head…through optical fibers…Raman spectrum at the fundamental wavelength of 785 nm was recorded”; Examiner notes that though BRULLOT does not recite explicit direction to “irradiate” a sample, it would be clear to one of ordinary skill the term “fundamental wavelength” is used for sample excitation (irradiation).)
collecting data responsive to the excitation beam using the spectroscopy device;
(BRULLOT teaches collecting Raman data in [0052] In one embodiment of the method of the present disclosure, the Raman spectrum of the superabsorbent polymer is collected in-line in a production process”, with detection range limits found in [0127]: “detection window ranged from 150 cm−1 to 1900 cm−1”)
and processing the data using a scientific instrument support apparatus according to claim 1.
(See above references as applied to Claim 1)
With respect to Claim 28, BRULLOT teaches:
A method for scientific instrument support,
(As above in parallel claim limitation presented in Claim 1, BRULLOT teaches a computer implemented data acquisition system and method, see [0115]-[0117], with data acquisition from a Raman spectrometer in [0119].)
comprising: managing and pre-processing data;
(As above, in parallel claim limitation presented in Claim 1, BRULLOT teaches data management and pre-processing of data, for example see FIG. 1 with [0013]: “normalization step as data pre-treatment”; and see [0045]-[0051] for disclosure of detailed techniques for pre-treatment of data; Examiner notes BRI applied to claim limitation language of “managing … data” to mean generally expression of a plan for how data is processed in a data analysis system, and “pre-processing…data” to mean generally analysis steps carried out prior to model development using data.)
training one or more models to provide a trained models;
(as above, BRULLOT teaches
providing a measure of the quality of the trained model;
(As above in parallel limitations in Claim 1, BRULLOT teaches training models, see [0043] and teaches development of trained models for data analysis, see [0037], and teaches validation and testing of models, see [0051]: “several models are made/trained and tested/validated with process samples”; and see Claim 2; Examiner notes again BRI as applied to claim limitation language of “second logic” as above)
However, BRULLOT is silent to the language of:
and providing a one or more hyperparameter of the trained model.
Nevertheless, SUI teaches:
and providing a one or more hyperparameter of the trained model.
(See as above, parallel claim limitation in Claim 1, SUI teaches use of hyperparameter optimization to train data analysis models, see FIG. 1 B) and Pg.2; SUI teaches hyperparameter selection, “a designed searcher will search for the optimal combination of hyperparameters so as to obtain the most suitable model for a specific task.”, and see Fig. 2 for process steps in using hyperparameter to train models.)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to modify BRULLOT to include providing a one or more hyperparameter of the trained model, such as that of SUI.
One of ordinary skill would be to modify BRULLOT to include providing a one or more hyperparameter of the trained model, as taught by SUI because it would be understood as a way to complete the use of the implementation of hyperparameter optimization as suggested in BRULLOT. One of ordinary skill would understand the known technique and value of using hyperparameter selection to train models for data analysis as a way to make analysis of large data sets more efficient and more accurate minimizing computational time and effort to arrive at a best model for given data. One of ordinary skill would be further motivated to use the machine learning/model training method as suggested by BRULLOT in combination with the detailed analysis techniques taught by SUI (and others, as included below in conclusion) because such processes are known improve the ability to achieve faster, more accurate and reliable quantitative and/or qualitative and/or identification information of a wide range of materials when used in the context of Raman spectroscopic data. One of ordinary skill would be motivated to use hyperparameters, as suggested by BRULLOT, and refined in the method of SUI with a reasonable expectation of success in ensuring models perform optimally and provide maximum predictive power, which would be advantageous particularly in an industrial setting.
With respect to Claim 29, BRULLOT, in view of SUI, teaches:
One or more non-transitory computer readable media having instructions thereon
(BRULLOT further teaches standard computational system components used for carrying out disclosed method, see [0114]: “computer program product may be used with or incorporated in a computer system that may be a standalone unit or include one or more remote terminals or devices in communication with a central computer via a network…computer or processor and related components described herein may be a portion of a local computer system or a remote computer or an on-line system or combinations thereof. Any database and the software product, i.e. the computer program product described herein may be stored in computer internal memory or in a non-transitory computer readable medium.”)
that, when executed by one or more processing devices of a scientific instrument support apparatus, cause the scientific instrument support apparatus to perform the method of claim 28.
(See as above, references as applied to Claim 28)
Claims 3, 9, 13, and 23- 25 are rejected under 35 U.S.C. 103(a) as being unpatentable over BRULLOT (US 20220003679 A1) in view of SUI (SUI, et al., "A deep learning model designed for Raman spectroscopy with a novel hyperparameter optimization method", Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 280 (2022) 121560 - Available online 25 June 2022), and further in view of BAUER (US 20220223230 A1).
With respect to Claim 3, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 2,
(See above, references as applied to Claim 2.)
However, BRULLOT as modified by SUI, as taught above, is silent to the language of:
wherein the different training samples include one or more of, a media variation, a processing parameter variation, a target material variation, a reactor variation, and a spectroscopic instrument variation.
Nevertheless, BAUER teaches:
wherein the different training samples include one or more of, a media variation, a processing parameter variation, a target material variation, a reactor variation, and a spectroscopic instrument variation.
(For context, see Abstract: “disclosure relates to automated systems and methods for quantitatively determining an unmasking status of a biological specimen subjected to an unmasking process…using a trained unmasking status estimation engine. In some embodiments, the trained unmasking status estimation engine comprises a machine learning algorithm based on a projection onto latent structure regression model. In some embodiments, the trained unmasking status estimation engine includes a neural network”, thus BAUER is in same technical area; BAUER teaches using variations in training samples, see [0014]: “estimation engine is trained using one or more training spectral data sets, wherein each training spectral data set comprises a plurality of training vibrational spectra derived from a plurality of differentially unmasked training tissue samples”; and see [0154]: “model is also trained to recognize the changes in these features for different types of tissues and/or for different types of molecules…algorithm takes the vibrational spectral…and creates a model that is used to determine which features (wavelengths) are most predictive of the response variable”)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify BRULLOT as modified by SUI, as taught above, to include wherein the different training samples include one or more of, a media variation, a processing parameter variation, a target material variation, a reactor variation, and a spectroscopic instrument variation, such as that of BAUER.
One of ordinary skill would be motivated to further modify BRULLOT as modified by SUI, as taught above, to include wherein the different training samples include one or more of, a media variation, a processing parameter variation, a target material variation, a reactor variation, and a spectroscopic instrument variation, as taught by BAUER because it would be understood as a way to develop more robust and generalizable models that would have a wide range of applications. Using a method as taught by BAUER combined with the method taught by BRULLOT as modified by SUI would be seen as a way to develop transferrable processes that could scale up and/or scale down as needed based on a particular application. One of ordinary skill would see reasonable expectation of success for improving process monitoring and control by combining the teaching of BRULLOT as modified by SUI with variations in training data as taught by BAUER, particularly in industrial settings where efficiency and reliability are valued.
With respect to Claim 9, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
However, BRULLOT, as modified by SUI, as taught above, is silent to the language of:
wherein the first logic accepts a problem type selected from a qualitative challenge or a quantitative challenge.
Nevertheless, BAUER teaches:
wherein the first logic accepts a problem type selected from a qualitative challenge or a quantitative challenge.
(BAUER teaches analysis method to product in both qualitative and quantitative results, see [0012]: “system and method which advantageously enables the quantitative and/or qualitative prediction of the unmasking status of a biological specimen”; Examiner notes interpretation of claim limitation language of “qualitative challenge or a quantitative challenge” to mean generally solving a problem resulting in a qualitative result or a quantitative result, respectively, analogous to reference teaching of “prediction of the unmasking status”.)
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify BRULLOT as modified by SUI, as taught above, to include wherein the first logic accepts a problem type selected from a qualitative challenge or a quantitative challenge, such as that of BAUER.
One of ordinary skill would be motivated to further modify BRULLOT as modified by SUI, as taught above, to include wherein the first logic accepts a problem type selected from a qualitative challenge or a quantitative challenge, as taught by BAUER because it would be understood as a way to develop more robust and generalizable models that would provide a more comprehensive understanding and insight with regard to data analysis, and yield broader decision and prediction making capacity. One of ordinary skill would see the holistic approach to data analysis with capacity to predict qualitative properties of a sample or system along with quantitative information as a way to provide a more robust contextual understanding of samples or systems under test. One of ordinary skill would understand the efficiency of qualitative data in making rapid decisions and the necessity of quantitative results for decisions such as process control or parameter regulation.
With respect to Claim 13, BRULLOT, in view of SUI, teaches:
The scientific instrument support apparatus according to claim 1,
(See above, references as applied to Claim 1.)
However, BRULLOT, as modified by SUI, as taught above, is silent to the language of:
wherein the model is input as a selection of different model types by a user to the second logic.
Nevertheless, BAUER teaches:
wherein the model is input as a selection of different model types by a user to the second logic.
(BAUER teaches user interaction with model throughout process, see [0227]: “provide for interaction with a user, embodiments of the subject matter described in this specification can be implemented on a computer having a display device”; and see [0228]: “subject matter described in this specification can be implemented in a computing system that includes a back-end component, e.g., as a data server, or that includes a middleware component…a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back-end, middleware, or front-end components”; Examiner notes interpretation of claim limitation language of “input…by a user” to be analogous to reference suggestion that user has input as all points in modeling process”
It would have been obvious to one of ordinary skill in the art before effective filing date of the claimed invention to further modify BRULLOT as modified by SUI, as taught above, to include wherein the model is input as a selection of different model types by a user to the second logic, such as that of BAUER.
One of ordinary skill would be motivated to further modify BRULLOT as modified by SUI, as taught above, to include wherein the model is input as a selection of different model types by a user to the second logic, as taught by BAUER because it would be understood as an option for including expertise of individuals who may possess valuable domain knowledge that could assist in model training of selection of best models. One of ordinary skill would see the advantage of combining the user involvement taught by BAUER with the system and method of BRULLOT, as modified by