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 .
Drawings
The drawings are objected to because:
Figures 1 and 3 contain text that is not clear. All drawings must be made by a process which will give them satisfactory reproduction characteristics. Every line, number, and letter must be durable, clean, black (except for color drawings), sufficiently dense and dark, and uniformly thick and well-defined. Numbers, letters, and reference characters should not be placed upon hatched or shaded surfaces (see 37 CFR 1.84(l) and 37 CFR 1.84(p)).
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(5) because they do not include the following reference signs mentioned in the description:
“203, “204”, “205”, 206”, and “207” in Figure 2
“505, “514”, “516”, and “517” in Figure 5.
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(5) because they include the following reference character not mentioned in the description:
“transfer learning system 112” in paragraph 0068, line 4.
Corrected drawing sheets in compliance with 37 CFR 1.121(d), or amendment to the specification to add the reference characters in the description in compliance with 37 CFR 1.121(b) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Specification
The disclosure is objected to because of the following informalities:
In paragraph 0006, line 7, “found ton only be” should read “found to only be”.
In paragraph 0071, line 10, “Similarity, the term” should read “Similarly, the term”.
In paragraph 0096, line 6, “member computing devices 401-404” should read “member computing devices 402-404”.
In paragraph 0099, line 17, “users, 512a through 502n” should read “users, 512a through 512n”.
Appropriate correction is required.
Claim Objections
Claims 1, 10, 16 and 20 are objected to because of the following informalities:
In claim 1, lines 6-7, “a model performance confusion matrix, model performance confusion matrix corresponding to” should read “a model performance confusion matrix, the model performance confusion matrix corresponding to”.
In claim 10, lines 1-2, “non-transitory computer-readable storage medium tangibly encoded without computer-executable instructions” should read “non-transitory computer-readable storage medium tangibly encoded with computer-executable instructions”.
In claim 10, lines 8-9, “a model performance confusion matrix, model performance confusion matrix corresponding to” should read “a model performance confusion matrix, the model performance confusion matrix corresponding to”.
In claim 16, lines 7-8, “a model performance confusion matrix, model performance confusion matrix corresponding to” should read “a model performance confusion matrix, the model performance confusion matrix corresponding to”.
In claim 20, line 1, “device of claim 16, further comprising” should read “device of claim 16, wherein the processor is further configured to”.
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1 – 20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites the limitation "generating, by the device, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" in lines 6-8. This limitation is indefinite because the meaning of “the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" is not clear. The specification recites, in paragraph 0038, lines 1-4, “In some embodiments, each model is tested against every SDS segment in the SDS dataset, generating a confusion matrix for each model using the four different aggregation methods (1) average (avg), (2) maximum (max), (3) vote (vote), and (4) vote average (votea) at 35 different threshold values starting at 0.05 up to 0.9 in 0.025 steps.”, disclosing generating a confusion matrix for each model in a set of neural network models by testing each model against every SDS segment in an SDS dataset using four aggregation methods at different threshold values. However, it is not clear if “the model performance confusion matrix corresponding to a set of models” refers to a confusion matrix for each model in a set of neural network models, and it is not clear if “a threshold based SDS analysis threshold” refers to one of the different threshold values used to generating a confusion matrix for each model in a set of neural network models.
Claim 1 further recites the limitation "performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions" in lines 12-15. This limitation is indefinite because the meaning of “the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” is not clear. The specification recites, in paragraph 0038, lines 3-6, “Each time, a new nofalse model is identified, it is fed to the test function. If the model increases the number of true positives or true negatives predictions the model is kept in the performance group, if it does not add any more true positives or true negatives, it is excluded.”, recites, in paragraph 0049, lines 5-6, “Accordingly, in some embodiments, the resulting model grouping may be designated as the "nofalse stack".”, recites, in paragraph 0051, lines 6-11, “Once this list of positive pairs and their performance are determined (or otherwise identified), the disclosed framework can select the best pairs as finalists, where those positive pairs are provided as the first models in a stack with each of the remaining positive predictors to find the best set of models for predicting positives. In some embodiments, a threshold is selected for the maximum number of false positives the model sets are allowed to predict. These models are saved as the "positive predictor stack".”, and recites, in paragraph 0052, lines 1-5, “In some embodiments, this process is repeated for negative predictors, attempting all pairs of models which got less than 10% of the negatives wrong to select the best pairs, then attempting those pairs with each of the remaining negative predictors to identify the 3-model stack that predicts negatives best without getting too many wrong (e.g., at or below a threshold value). These models are grouped as the "negative predictor stack".”. However, it is not clear if “the type of output corresponding to a positive, negative and nofalse predictions” refers to the prediction criteria used to group models into the "nofalse stack", the "positive predictor stack", and the "negative predictor stack". Also, it is not clear if “a set of configuration models” refers to the set of neural network models used to generate the model performance confusion matrix, the neural network models that produce a quantity of false outputs at or below a threshold level, or a different set of models.
Claim 1 further recites the limitation “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” is lines 16-17. This limitation is indefinite because the meaning of “nofalse predictions” is not clear. The specification recites, in paragraph 0047, lines 1-6, “According to some embodiments, use of a confusion matrix (e.g., allresults table) for a given model can be effectuated to select the optimum "nofalse" configuration(s) for each model based on the threshold values tested. The nofalse configuration is the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on the test set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives.”, disclosing a nofalse configuration for a model being the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on a test data set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives. However, it is not clear how a nofalse prediction corresponds to a nofalse configuration. Also, the difference between a “model performance grouping” and a “stack” is not clear.
Claim 1 is also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The specification recites, in paragraph 0053, lines 1-5, “Accordingly, the grouped stacks, supra, can be assembled into a nearly complete oracle. In some embodiments, the nofalse stack is on top, below that go the positive predictors, then the negative predictors. In some embodiments, the oracle can be permitted to be inconclusive ('I don't know'), whereby according to such embodiments, the definition of oracle can be considered complete.”, recites, in paragraph 0055, lines 1-6, “In some embodiments and, optionally, in combination of any embodiment described above or below, once the no false stack is identified, a definition of the model grouping is created in a file, database, array, memory, and the like, or some combination thereof. This is called the oracle definition, provide a listing of model, the structure of the model grouping, order they should be used as well as model parameters and thresholds, and aggregation methods that should be used for each model to predict positives and/or negatives.”, recites, in paragraph 0056, lines 1-5, “According to some embodiments, in order to predict an SDS, the system starts by obtaining the predictions from the first model pairing for all the SDS segments (a single cough), then aggregating those predictions using the method specified by posmeth at the threshold specified by posthresh for that model. If that results in a positive prediction, then that SDS is predicted as positive and the system returns a positive result and processing is done.”, and recites, in paragraph 0057, lines 1-6, “If the sample is not predicted as positive, the system evaluates negmeth at negthresh in the model grouping. If negative then the sample is predicted as negative, a negative result is output/returned, and processing is done. If this model does not predict the sample as either positive or negative, then that constitutes an inconclusive result ('I don't know'), and the system processes to the next model. This is repeated until a prediction is reached or the model list is exhausted with inconclusive results.”. The step of defining the oracle and the step of using the oracle to determine an SDS classification for the audio file are essential to the invention as described in the specification of performing a signal data signature classification for an audio file.
Claims 2 – 9 are also rejected as they depend from claim 1 and thus recite the limitations of claim 1, and do not resolve the indefinite language or omission of essential steps from claim 1.
Claim 3 also recites the limitation "the model configuration for each model in the stack" in lines 3-4. There is insufficient antecedent basis for this limitation in the claim.
Claim 10 recites the limitation "generating, by the device, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" in lines 8-10. This limitation is indefinite because the meaning of “the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" is not clear. The specification recites, in paragraph 0038, lines 1-4, “In some embodiments, each model is tested against every SDS segment in the SDS dataset, generating a confusion matrix for each model using the four different aggregation methods (1) average (avg), (2) maximum (max), (3) vote (vote), and (4) vote average (votea) at 35 different threshold values starting at 0.05 up to 0.9 in 0.025 steps.”, disclosing generating a confusion matrix for each model in a set of neural network models by testing each model against every SDS segment in an SDS dataset using four aggregation methods at different threshold values. However, it is not clear if “the model performance confusion matrix corresponding to a set of models” refers to a confusion matrix for each model in a set of neural network models, and it is not clear if “a threshold based SDS analysis threshold” refers to one of the different threshold values used to generating a confusion matrix for each model in a set of neural network models.
Claim 10 further recites the limitation "performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions" in lines 14-17. This limitation is indefinite because the meaning of “the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” is not clear. The specification recites, in paragraph 0038, lines 3-6, “Each time, a new nofalse model is identified, it is fed to the test function. If the model increases the number of true positives or true negatives predictions the model is kept in the performance group, if it does not add any more true positives or true negatives, it is excluded.”, recites, in paragraph 0049, lines 5-6, “Accordingly, in some embodiments, the resulting model grouping may be designated as the "nofalse stack".”, recites, in paragraph 0051, lines 6-11, “Once this list of positive pairs and their performance are determined (or otherwise identified), the disclosed framework can select the best pairs as finalists, where those positive pairs are provided as the first models in a stack with each of the remaining positive predictors to find the best set of models for predicting positives. In some embodiments, a threshold is selected for the maximum number of false positives the model sets are allowed to predict. These models are saved as the "positive predictor stack".”, and recites, in paragraph 0052, lines 1-5, “In some embodiments, this process is repeated for negative predictors, attempting all pairs of models which got less than 10% of the negatives wrong to select the best pairs, then attempting those pairs with each of the remaining negative predictors to identify the 3-model stack that predicts negatives best without getting too many wrong (e.g., at or below a threshold value). These models are grouped as the "negative predictor stack".”. However, it is not clear if “the type of output corresponding to a positive, negative and nofalse predictions” refers to the prediction criteria used to group models into the "nofalse stack", the "positive predictor stack", and the "negative predictor stack". Also, it is not clear if “a set of configuration models” refers to the set of neural network models used to generate the model performance confusion matrix, the neural network models that produce a quantity of false outputs at or below a threshold level, or a different set of models.
Claim 10 further recites the limitation “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” is lines 18-19. This limitation is indefinite because the meaning of “nofalse predictions” is not clear. The specification recites, in paragraph 0047, lines 1-6, “According to some embodiments, use of a confusion matrix (e.g., allresults table) for a given model can be effectuated to select the optimum "nofalse" configuration(s) for each model based on the threshold values tested. The nofalse configuration is the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on the test set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives.”, disclosing a nofalse configuration for a model being the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on a test data set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives. However, it is not clear how a nofalse prediction corresponds to a nofalse configuration. Also, the difference between a “model performance grouping” and a “stack” is not clear.
Claim 10 is also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The specification recites, in paragraph 0053, lines 1-5, “Accordingly, the grouped stacks, supra, can be assembled into a nearly complete oracle. In some embodiments, the nofalse stack is on top, below that go the positive predictors, then the negative predictors. In some embodiments, the oracle can be permitted to be inconclusive ('I don't know'), whereby according to such embodiments, the definition of oracle can be considered complete.”, recites, in paragraph 0055, lines 1-6, “In some embodiments and, optionally, in combination of any embodiment described above or below, once the no false stack is identified, a definition of the model grouping is created in a file, database, array, memory, and the like, or some combination thereof. This is called the oracle definition, provide a listing of model, the structure of the model grouping, order they should be used as well as model parameters and thresholds, and aggregation methods that should be used for each model to predict positives and/or negatives.”, recites, in paragraph 0056, lines 1-5, “According to some embodiments, in order to predict an SDS, the system starts by obtaining the predictions from the first model pairing for all the SDS segments (a single cough), then aggregating those predictions using the method specified by posmeth at the threshold specified by posthresh for that model. If that results in a positive prediction, then that SDS is predicted as positive and the system returns a positive result and processing is done.”, and recites, in paragraph 0057, lines 1-6, “If the sample is not predicted as positive, the system evaluates negmeth at negthresh in the model grouping. If negative then the sample is predicted as negative, a negative result is output/returned, and processing is done. If this model does not predict the sample as either positive or negative, then that constitutes an inconclusive result ('I don't know'), and the system processes to the next model. This is repeated until a prediction is reached or the model list is exhausted with inconclusive results.”. The step of defining the oracle and the step of using the oracle to determine an SDS classification for the audio file are essential to the invention as described in the specification of performing a signal data signature classification for an audio file.
Claims 11 – 15 are also rejected as they depend from claim 10 and thus recite the limitations of claim 10, and do not resolve the indefinite language or omission of essential steps from claim 10.
Claim 12 also recites the limitation "the model configuration for each model in the stack" in lines 4-5. There is insufficient antecedent basis for this limitation in the claim.
Claim 16 recites the limitation "generating, by the device, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" in lines 7-9. This limitation is indefinite because the meaning of “the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold" is not clear. The specification recites, in paragraph 0038, lines 1-4, “In some embodiments, each model is tested against every SDS segment in the SDS dataset, generating a confusion matrix for each model using the four different aggregation methods (1) average (avg), (2) maximum (max), (3) vote (vote), and (4) vote average (votea) at 35 different threshold values starting at 0.05 up to 0.9 in 0.025 steps.”, disclosing generating a confusion matrix for each model in a set of neural network models by testing each model against every SDS segment in an SDS dataset using four aggregation methods at different threshold values. However, it is not clear if “the model performance confusion matrix corresponding to a set of models” refers to a confusion matrix for each model in a set of neural network models, and it is not clear if “a threshold based SDS analysis threshold” refers to one of the different threshold values used to generating a confusion matrix for each model in a set of neural network models.
Claim 16 further recites the limitation "performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions" in lines 13-16. This limitation is indefinite because the meaning of “the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” is not clear. The specification recites, in paragraph 0038, lines 3-6, “Each time, a new nofalse model is identified, it is fed to the test function. If the model increases the number of true positives or true negatives predictions the model is kept in the performance group, if it does not add any more true positives or true negatives, it is excluded.”, recites, in paragraph 0049, lines 5-6, “Accordingly, in some embodiments, the resulting model grouping may be designated as the "nofalse stack".”, recites, in paragraph 0051, lines 6-11, “Once this list of positive pairs and their performance are determined (or otherwise identified), the disclosed framework can select the best pairs as finalists, where those positive pairs are provided as the first models in a stack with each of the remaining positive predictors to find the best set of models for predicting positives. In some embodiments, a threshold is selected for the maximum number of false positives the model sets are allowed to predict. These models are saved as the "positive predictor stack".”, and recites, in paragraph 0052, lines 1-5, “In some embodiments, this process is repeated for negative predictors, attempting all pairs of models which got less than 10% of the negatives wrong to select the best pairs, then attempting those pairs with each of the remaining negative predictors to identify the 3-model stack that predicts negatives best without getting too many wrong (e.g., at or below a threshold value). These models are grouped as the "negative predictor stack".”. However, it is not clear if “the type of output corresponding to a positive, negative and nofalse predictions” refers to the prediction criteria used to group models into the "nofalse stack", the "positive predictor stack", and the "negative predictor stack". Also, it is not clear if “a set of configuration models” refers to the set of neural network models used to generate the model performance confusion matrix, the neural network models that produce a quantity of false outputs at or below a threshold level, or a different set of models.
Claim 16 further recites the limitation “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” is lines 17-18. This limitation is indefinite because the meaning of “nofalse predictions” is not clear. The specification recites, in paragraph 0047, lines 1-6, “According to some embodiments, use of a confusion matrix (e.g., allresults table) for a given model can be effectuated to select the optimum "nofalse" configuration(s) for each model based on the threshold values tested. The nofalse configuration is the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on the test set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives.”, disclosing a nofalse configuration for a model being the positive threshold and method that provides the greatest number of true positives with zero false positives from that model on a test data set, as well as the negative threshold and method that provides the greatest number of true negatives with no false negatives. However, it is not clear how a nofalse prediction corresponds to a nofalse configuration. Also, the difference between a “model performance grouping” and a “stack” is not clear.
Claim 16 is also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The specification recites, in paragraph 0053, lines 1-5, “Accordingly, the grouped stacks, supra, can be assembled into a nearly complete oracle. In some embodiments, the nofalse stack is on top, below that go the positive predictors, then the negative predictors. In some embodiments, the oracle can be permitted to be inconclusive ('I don't know'), whereby according to such embodiments, the definition of oracle can be considered complete.”, recites, in paragraph 0055, lines 1-6, “In some embodiments and, optionally, in combination of any embodiment described above or below, once the no false stack is identified, a definition of the model grouping is created in a file, database, array, memory, and the like, or some combination thereof. This is called the oracle definition, provide a listing of model, the structure of the model grouping, order they should be used as well as model parameters and thresholds, and aggregation methods that should be used for each model to predict positives and/or negatives.”, recites, in paragraph 0056, lines 1-5, “According to some embodiments, in order to predict an SDS, the system starts by obtaining the predictions from the first model pairing for all the SDS segments (a single cough), then aggregating those predictions using the method specified by posmeth at the threshold specified by posthresh for that model. If that results in a positive prediction, then that SDS is predicted as positive and the system returns a positive result and processing is done.”, and recites, in paragraph 0057, lines 1-6, “If the sample is not predicted as positive, the system evaluates negmeth at negthresh in the model grouping. If negative then the sample is predicted as negative, a negative result is output/returned, and processing is done. If this model does not predict the sample as either positive or negative, then that constitutes an inconclusive result ('I don't know'), and the system processes to the next model. This is repeated until a prediction is reached or the model list is exhausted with inconclusive results.”. The step of defining the oracle and the step of using the oracle to determine an SDS classification for the audio file are essential to the invention as described in the specification of performing a signal data signature classification for an audio file.
Claims 17 – 20 are also rejected as they depend from claim 16 and thus recite the limitations of claim 16, and do not resolve the indefinite language or omission of essential steps from claim 16.
Claim 18 also recites the limitation "the model configuration for each model in the stack" in lines 3-4. There is insufficient antecedent basis for this limitation in the claim.
Allowable Subject Matter
Claims 1 – 20 would be allowable if rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action.
The following is a statement of reasons for the indication of allowable subject matter:
The primary reason claim 1 would be allowable if rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action, is the inclusion of the limitations “performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” in combination with the limitations to receive a request for classification of an audio file, the audio file comprising audio content, analyze the audio file and determining a signal data signature (SDS) for the audio file, generate, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold, perform a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level, perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, generate, based on the model performance grouping, a set of stacks, assemble an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database, and determining and outputting an SDS classification for the audio file.
Ramirez et al. (US Patent Application Publication No. 2022/0215248), hereinafter Ramirez, discloses a method comprising:
receiving, by a device, a request for classification of an audio file, the audio file comprising audio content (Paragraph 0041, lines 1-5, "Embodiments of the present disclosure are directed to the signal data signature detection system 100 whereby a signal data recording (the input 101) is provided by an individual or individuals(s) or system into a computer hardware");
analyzing, by the device, the audio file, and determining a signal data signature (SDS) for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries.");
a set of models that correspond to a threshold based SDS analysis threshold (Paragraph 0043, lines 8-16, "The signal data signature classification system 111 executes a signal data signature classifier system 112 on a processor 105 such that the paired training dataset is used to train machine learning (ML) models 113 that generate boundaries within the dataset 114 whereby the boundaries inform the scope and datasets of target model(s) 121 and the source model 116, such that knowledge is transferred 117 from the source model 116 to the target model(s) 121.");
and determining and outputting, by the device, an SDS classification for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries."; Paragraph 0048, lines 1-12, "In one or more embodiments of the signal data signature detection system 100 the output 118 includes a strongly labeled signal data signature recording and identification of signal data signature type. An example would be signal data signature sample from a patient which would include: 1) a label of the identified signal data signature type, 2) or flag that tells the user that a signal data signature was not detected. The output 118 of signal data signature type or message that a signal data signature was not detected will be delivered to an end user via a display medium such as but not limited to a display screen 119 (e.g., tablet, mobile phone, computer screen) and/or paper 120.").
Tandecki et al. (US Patent No. 11,138,477), hereinafter Tandecki, teaches:
generating, by the device, based on a set of neural network models, a model performance confusion matrix, model performance confusion matrix corresponding to a set of models that correspond to a threshold based analysis threshold (Column 7, line 64 - Column 8, line 5, 'The classification models may determine a confidence level that estimates a certainty of the proposed classification based on previous classifications performed. The confidence level (or percentage) may be based on training data. In some embodiments, with columnar data, the classification modules may generate a proposed classification for every data entity in a data column, but these proposed classifications may be combined to generate an overall classification for the entire column."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector.");
performing, by the device, a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level (Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 16, lines 20-25, "A false positive may differ from a false negative. In some embodiments, the system may end up with returning only some values above certain threshold, which may return only those which are considered the most appropriate. The system may go for as low as possible false positive rate even if the system ends up with returning no matching classes.");
performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model (Column 4, lines 32-50, " The present embodiments relate to classifying portions of data using aggregated classification information received from multiple classification modules. Multiple classification modules may use various techniques (e.g., dictionaries, regular expression (Regex) pattern matching, a neural network) to inspect received data and determine proposed classification(s) for the data and a confidence level in each proposed classification. The varying techniques utilized by the classification modules may provide one or more proposed classifications with differing confidences in each proposed classification. Additionally, an aggregation module (or “aggregator”) can receive and combine the proposed classifications and confidences and inspect the aggregated information to determine a final classification for the data that represents a classification with a greater degree of accuracy. Based on the final classification determined by the aggregator, a further action (e.g., storing the data, sending the data to a client device, encrypting the data) can be performed."; Column 16, line 64 - Column 17, line 4, "In some embodiments, the aggregator may make a flexible decision based on the results. It may be possible to separate a “leader group” from all of results. A leader group may be a group of top results which are relatively close to each other and relatively far from the rest of results. Accordingly, determining and finding a “leader group” may include clustering, calculating standard deviation for and between groups, etc."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”").
Soppin et al. (US Patent No. 11,526,814), hereinafter Soppin, teaches:
generating, by the device, based on the model performance grouping, a set of stacks (Column 1, line 61 - Column 2, line 4, "In one embodiment, a method for building an ensemble model is disclosed. In one example, the method may include creating a plurality of clusters, each including a set of predictive models forming the ensemble model. The method may further include initializing, for each of the plurality of clusters, each of the set of predictive models with random values for a set of first parameters to obtain a first accuracy score. The method may further include categorizing, for each of the plurality of clusters, each of the set of predictive models into a first associated category based on the first accuracy score.");
assembling, by the device, an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database (Column 5, lines 42-48, "Further, the initialization unit 201 may send the set of predictive models initialized with the set of first parameters to the decision unit 202 to initiate training process. In some embodiments, the initialization unit 201 may be configured to send the model ID for each of the set of predictive models, the cluster ID for each of the plurality of clusters, and the set of first parameters to the data repository 206.").
However, Ramirez, Tandecki, and Soppin, individually or in combination, do not disclose the limitations “performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions”.
Claims 2 – 9 depend from claim 1 and thus recite the limitations of claim 1, and would be allowable if claim 1 was rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action.
The primary reason claim 10 would be allowable if rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action, is the inclusion of the limitations “performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” in combination with the limitations to receive a request for classification of an audio file, the audio file comprising audio content, analyze the audio file and determining a signal data signature (SDS) for the audio file, generate, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold, perform a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level, perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, generate, based on the model performance grouping, a set of stacks, assemble an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database, and determining and outputting an SDS classification for the audio file.
Ramirez discloses a non-transitory computer-readable storage medium tangibly encoded without computer-executable instructions (Paragraph 0043, lines 1-5, "In some embodiments, the data sources 108 and the signal data signature recording input 101 are stored in memory or a memory unit 104 and passed to a software 109 such as computer program or computer programs that executes the instruction set on a processor 105."), that when executed by a device, perform a method comprising:
receiving, by the device, a request for classification of an audio file, the audio file comprising audio content (Paragraph 0041, lines 1-5, "Embodiments of the present disclosure are directed to the signal data signature detection system 100 whereby a signal data recording (the input 101) is provided by an individual or individuals(s) or system into a computer hardware");
analyzing, by the device, the audio file, and determining a signal data signature (SDS) for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries.");
a set of models that correspond to a threshold based SDS analysis threshold (Paragraph 0043, lines 8-16, "The signal data signature classification system 111 executes a signal data signature classifier system 112 on a processor 105 such that the paired training dataset is used to train machine learning (ML) models 113 that generate boundaries within the dataset 114 whereby the boundaries inform the scope and datasets of target model(s) 121 and the source model 116, such that knowledge is transferred 117 from the source model 116 to the target model(s) 121.");
and determining and outputting, by the device, an SDS classification for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries."; Paragraph 0048, lines 1-12, "In one or more embodiments of the signal data signature detection system 100 the output 118 includes a strongly labeled signal data signature recording and identification of signal data signature type. An example would be signal data signature sample from a patient which would include: 1) a label of the identified signal data signature type, 2) or flag that tells the user that a signal data signature was not detected. The output 118 of signal data signature type or message that a signal data signature was not detected will be delivered to an end user via a display medium such as but not limited to a display screen 119 (e.g., tablet, mobile phone, computer screen) and/or paper 120.").
Tandecki teaches:
generating, by the device, based on a set of neural network models, a model performance confusion matrix, model performance confusion matrix corresponding to a set of models that correspond to a threshold based analysis threshold (Column 7, line 64 - Column 8, line 5, 'The classification models may determine a confidence level that estimates a certainty of the proposed classification based on previous classifications performed. The confidence level (or percentage) may be based on training data. In some embodiments, with columnar data, the classification modules may generate a proposed classification for every data entity in a data column, but these proposed classifications may be combined to generate an overall classification for the entire column."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector.");
performing, by the device, a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level (Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 16, lines 20-25, "A false positive may differ from a false negative. In some embodiments, the system may end up with returning only some values above certain threshold, which may return only those which are considered the most appropriate. The system may go for as low as possible false positive rate even if the system ends up with returning no matching classes.");
performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model (Column 4, lines 32-50, " The present embodiments relate to classifying portions of data using aggregated classification information received from multiple classification modules. Multiple classification modules may use various techniques (e.g., dictionaries, regular expression (Regex) pattern matching, a neural network) to inspect received data and determine proposed classification(s) for the data and a confidence level in each proposed classification. The varying techniques utilized by the classification modules may provide one or more proposed classifications with differing confidences in each proposed classification. Additionally, an aggregation module (or “aggregator”) can receive and combine the proposed classifications and confidences and inspect the aggregated information to determine a final classification for the data that represents a classification with a greater degree of accuracy. Based on the final classification determined by the aggregator, a further action (e.g., storing the data, sending the data to a client device, encrypting the data) can be performed."; Column 16, line 64 - Column 17, line 4, "In some embodiments, the aggregator may make a flexible decision based on the results. It may be possible to separate a “leader group” from all of results. A leader group may be a group of top results which are relatively close to each other and relatively far from the rest of results. Accordingly, determining and finding a “leader group” may include clustering, calculating standard deviation for and between groups, etc."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”").
Soppin teaches:
generating, by the device, based on the model performance grouping, a set of stacks (Column 1, line 61 - Column 2, line 4, "In one embodiment, a method for building an ensemble model is disclosed. In one example, the method may include creating a plurality of clusters, each including a set of predictive models forming the ensemble model. The method may further include initializing, for each of the plurality of clusters, each of the set of predictive models with random values for a set of first parameters to obtain a first accuracy score. The method may further include categorizing, for each of the plurality of clusters, each of the set of predictive models into a first associated category based on the first accuracy score.");
assembling, by the device, an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database (Column 5, lines 42-48, "Further, the initialization unit 201 may send the set of predictive models initialized with the set of first parameters to the decision unit 202 to initiate training process. In some embodiments, the initialization unit 201 may be configured to send the model ID for each of the set of predictive models, the cluster ID for each of the plurality of clusters, and the set of first parameters to the data repository 206.").
However, Ramirez, Tandecki, and Soppin, individually or in combination, do not disclose the limitations “performing, by the device, a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generating, by the device, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions”.
Claims 11 – 15 depend from claim 10 and thus recite the limitations of claim 10, and would be allowable if claim 10 was rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action.
The primary reason claim 16 would be allowable if rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action, is the inclusion of the limitations “perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generate, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions” in combination with the limitations to receive a request for classification of an audio file, the audio file comprising audio content, analyze the audio file and determining a signal data signature (SDS) for the audio file, generate, based on a set of neural network models, a model performance confusion matrix, the model performance confusion matrix corresponding to a set of models that correspond to a threshold based SDS analysis threshold, perform a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level, perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, generate, based on the model performance grouping, a set of stacks, assemble an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database, and determining and outputting an SDS classification for the audio file.
Ramirez discloses a device comprising: a processor (Paragraph 0043, lines 1-5, "In some embodiments, the data sources 108 and the signal data signature recording input 101 are stored in memory or a memory unit 104 and passed to a software 109 such as computer program or computer programs that executes the instruction set on a processor 105.") configured to:
receive a request for classification of an audio file, the audio file comprising audio content (Paragraph 0041, lines 1-5, "Embodiments of the present disclosure are directed to the signal data signature detection system 100 whereby a signal data recording (the input 101) is provided by an individual or individuals(s) or system into a computer hardware");
analyze the audio file, and determine a signal data signature (SDS) for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries.");
a set of models that correspond to a threshold based SDS analysis threshold (Paragraph 0043, lines 8-16, "The signal data signature classification system 111 executes a signal data signature classifier system 112 on a processor 105 such that the paired training dataset is used to train machine learning (ML) models 113 that generate boundaries within the dataset 114 whereby the boundaries inform the scope and datasets of target model(s) 121 and the source model 116, such that knowledge is transferred 117 from the source model 116 to the target model(s) 121.");
and determine and output an SDS classification for the audio file (Paragraph 0025, lines 1-7, 'The present disclosure relates generally to machine learning classifiers. Embodiments of the present disclosure include signal data signature detection, signal data signature classification, utilizing-a strategic machine learning as a method and system for use of federated data, machine learning and swarm learning for a derived strategic blueprint facilitating machine learning across data boundaries."; Paragraph 0048, lines 1-12, "In one or more embodiments of the signal data signature detection system 100 the output 118 includes a strongly labeled signal data signature recording and identification of signal data signature type. An example would be signal data signature sample from a patient which would include: 1) a label of the identified signal data signature type, 2) or flag that tells the user that a signal data signature was not detected. The output 118 of signal data signature type or message that a signal data signature was not detected will be delivered to an end user via a display medium such as but not limited to a display screen 119 (e.g., tablet, mobile phone, computer screen) and/or paper 120.").
Tandecki teaches:
generate, based on a set of neural network models, a model performance confusion matrix, model performance confusion matrix corresponding to a set of models that correspond to a threshold based analysis threshold (Column 7, line 64 - Column 8, line 5, 'The classification models may determine a confidence level that estimates a certainty of the proposed classification based on previous classifications performed. The confidence level (or percentage) may be based on training data. In some embodiments, with columnar data, the classification modules may generate a proposed classification for every data entity in a data column, but these proposed classifications may be combined to generate an overall classification for the entire column."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector.");
perform a performance evaluation based on the model performance confusion matrix, the performance evaluation comprising determining neural network models that produce a quantity of false outputs at or below a threshold level (Column 15, lines 25-31, "Utilizing multiple approaches may minimize error per each classification using coefficients. A threshold may be chosen, and coefficients may be based on a confusion matrix. A long vector may be included as an input for a neural net to minimize error. For each classification, a learning set may be generated by transforming and projecting a classifier vector and expected vector."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”"; Column 16, lines 20-25, "A false positive may differ from a false negative. In some embodiments, the system may end up with returning only some values above certain threshold, which may return only those which are considered the most appropriate. The system may go for as low as possible false positive rate even if the system ends up with returning no matching classes.");
perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model (Column 4, lines 32-50, " The present embodiments relate to classifying portions of data using aggregated classification information received from multiple classification modules. Multiple classification modules may use various techniques (e.g., dictionaries, regular expression (Regex) pattern matching, a neural network) to inspect received data and determine proposed classification(s) for the data and a confidence level in each proposed classification. The varying techniques utilized by the classification modules may provide one or more proposed classifications with differing confidences in each proposed classification. Additionally, an aggregation module (or “aggregator”) can receive and combine the proposed classifications and confidences and inspect the aggregated information to determine a final classification for the data that represents a classification with a greater degree of accuracy. Based on the final classification determined by the aggregator, a further action (e.g., storing the data, sending the data to a client device, encrypting the data) can be performed."; Column 16, line 64 - Column 17, line 4, "In some embodiments, the aggregator may make a flexible decision based on the results. It may be possible to separate a “leader group” from all of results. A leader group may be a group of top results which are relatively close to each other and relatively far from the rest of results. Accordingly, determining and finding a “leader group” may include clustering, calculating standard deviation for and between groups, etc."; Column 16, lines 6-14, "Evaluation metrics may include a confusion matrix that is indicative of how often the aggregator identifies the correct and incorrect classification in its final classification. A confusion matrix may include Boolean values (true-false) that divide the result into 4 types: True Positive (TP)—properly classified “true” class, True Negative (TN)—properly classified “false” class, False Negative (FN)—“true” class classified as “false”, and False Positive (FP)—“false” class classified as “true.”").
Soppin teaches:
generate, based on the model performance grouping, a set of stacks (Column 1, line 61 - Column 2, line 4, "In one embodiment, a method for building an ensemble model is disclosed. In one example, the method may include creating a plurality of clusters, each including a set of predictive models forming the ensemble model. The method may further include initializing, for each of the plurality of clusters, each of the set of predictive models with random values for a set of first parameters to obtain a first accuracy score. The method may further include categorizing, for each of the plurality of clusters, each of the set of predictive models into a first associated category based on the first accuracy score.");
assemble an oracle data structure based on the generated set of stacks, the assembly of the oracle data structure comprising storage in a database (Column 5, lines 42-48, "Further, the initialization unit 201 may send the set of predictive models initialized with the set of first parameters to the decision unit 202 to initiate training process. In some embodiments, the initialization unit 201 may be configured to send the model ID for each of the set of predictive models, the cluster ID for each of the plurality of clusters, and the set of first parameters to the data repository 206.").
However, Ramirez, Tandecki, and Soppin, individually or in combination, do not disclose the limitations “perform a model performance grouping, the model performance grouping comprising organizing a set of configuration models based on a type of output from each configuration model, the type of output corresponding to a positive, negative and nofalse predictions” and “generate, based on the model performance grouping, a set of stacks, the set of stacks corresponding to the positive, negative and nofalse predictions”.
Claims 17 – 20 depend from claim 16 and thus recite the limitations of claim 16, and would be allowable if claim 16 was rewritten or amended to overcome the rejections under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), 2nd paragraph, set forth in this Office action.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Aljbawi et al. ("Developing a Multi-variate Prediction Model for the Detection of COVID-19 From Crowd-sourced Respiratory Voice Data") teaches a deep learning model identifying COVID-19 using voice recordings data provided by the general population.
Hemdan et al. ("CR19: A framework for preliminary detection of COVID-19 in cough audio signals using machine learning algorithms for automated medical diagnosis applications") teaches a framework for efficiently COVID-19 detection and diagnosis using hybrid machine learning algorithms with genetic algorithms from cough audio signals.
Anupam et al. ("Preliminary Diagnosis of COVID-19 Based on Cough Sounds Using Machine Learning Algorithms") teaches a methos for COVID-19 classification of cough sounds based on machine learning.
Clertant et al. ("Interpretable Cascade Classifiers with Abstention") teaches a framework for handling a general dynamic diagnostic protocol setting with cost-sensitive heterogeneous cascading systems.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to James Boggs whose telephone number is (571)272-2968. The examiner can normally be reached M-F 8:00 AM - 5:00 PM.
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, Daniel Washburn can be reached at (571)272-5551. 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.
/JAMES BOGGS/Examiner, Art Unit 2657
Prosecution Insights