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 .
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
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.
Claim 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.
Regarding claim 1 and 7 recite …. carrying out rainstorm early warning, and carrying out rainstorm prevention and control of transmission lines in advance. It is unclear from the recited limitation regarding “… carrying out rainstorm prevention and control of transmission lines in advance” as what steps or activities to be considered as carrying out rainstorm prevention as rainstorm are natural environmental activities which cannot be prevented. Similarly, the control of transmission lines in advance does not provide a clear meaning as what activity or activities can be considered to control of transmission lines in advance. Therefore, the claims 1 and 7 and dependent claims 2-6, 8-20 are rejected under 35 U.S.C 112 (b) as failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Claim 2, 11 and 16 recites the limitation "… the new parallel network close… the old parallel network" in lines 2 to 3. There is insufficient antecedent basis for this limitation in the claim. It is unclear whether the new parallel network and the old parallel network are same or different from claim 1.
Therefore, the claims 2, 11 and 16 and the dependent claims 10 is also rejected under 35 U.S.C 112 (b) as failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
The term “close” in claim 2, 4, 10, 11, 13, 16, 18 is a relative term which renders the claim indefinite. The term “close” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
Therefore, the claims 2, 4, 10, 11, 13, 16, 18 and their dependent claims 10, 5, 14 and 19 are also rejected under 35 U.S.C 112 (b) as failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Regarding claim 4, 10, 13 and 18 recite wherein the selectively constraining the drift of the low- frequency components of the multi-scale features by using the multi-scale feature frequency domain distillation technology, and the memorizing the knowledge learned by the parallel convolutional neural networks in the old tasks are specifically as follows:” It is unclear which of the limitations that follow belong to which step (constraining or memorizing).
Additionally, it is unclear how the recited limitations read on either selectively constraining or memorizing since constraining or memorizing is not recited in the function.
Therefore, the claims 4, 10, 13 and 18 and their dependent claims 5, 14, 19 are also rejected under 35 U.S.C 112 (b) as failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Regarding claims 7, 8, 16-20 recite A/The weather early device of claim… it is unclear to the Examiner what device to be considered a weather early device.
Therefore, the claims 7, 8, 16-20 are rejected under 35 U.S.C 112 (b) as failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention.
Dependent claims 2-6, 8-20 inherit the deficiency of the independent claims are likewise rejected under 35 U.S.C 112 (b).
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 1-20 are rejected under 35 U.S.C 101 because the claimed invention is directed to judicial exception (i.e., a law of nature, natural phenomenon, or an abstract idea) without significantly more.
Specifically, claim 1 recites:
A weather warning method based on incremental learning, comprising the steps:
down-sampling marine data to obtain multi-scale marine data, and dividing the multi-scale data into a plurality of task sequences bounded by a preset year;
inputting the task sequences into parallel convolutional neural networks in a data flow form, and extracting multi-scale features through supervised representation learning;
selectively constraining, by a multi-scale feature frequency domain distillation technology, drift of low-frequency components of the multi-scale features based on incremental training, and memorizing knowledge learned by the parallel convolutional neural networks in old tasks;
adaptively learning different fusion parameters according to different time spans of the input multi-scale data by using a multi-scale feature adaptive fusion technology, so as to enhance the ability to learn new tasks; and
outputting a Nino3.4 index reflecting a change rule of El Nino through fully connected layers according to the adaptively fused features, establishing a mapping function of an extreme rainfall probability r based on the Nino3.4 index, and in response to predicting that the value r goes beyond a threshold value k, carrying out rainstorm early warning, and
carrying out rainstorm prevention and control of transmission lines in advance.
The claim limitations in the abstract idea have been highlighted in bold above.
Under the step 1 of the eligibility analysis, it is determined whether the claims are drawn to a statutory category by considering whether the claimed subject matter fall within the four statutory categories of patentable subject matter identified by 35 U.S.C 101: process, machine, manufacture, or composition of matter. The above claim is considered to be in the statutory category of (machine).
Under the step 2A, prong one, it is considered whether the claim recites a judicial exception (abstract idea). In the above claim, the highlighted portion constitutes an abstract idea because, under a broadest reasonable interpretation, it recites limitations that fall into/recite an abstract idea exceptions. Specifically, under the 2019 Revised Patent Subject Matter Eligibility Guidance, it falls into groupings of subject matter when recited as such in a claim limitation, that cover mathematical concepts (mathematical relationships, mathematical formulas or equations, mathematical calculations) and mental process – concepts performed in the human mind including an observation, evaluation, judgement, and/or opinion.
For example, a step of down-sampling marine data to obtain multi-scale marine data, and dividing the multi-scale data into a plurality of task sequences bounded by a preset year (is considered to be a mathematical step);
inputting the task sequences into parallel convolutional neural networks in a data flow form, and extracting multi-scale features through supervised representation learning (is considered to be a mathematical step);
selectively constraining, by a multi-scale feature frequency domain distillation technology, drift of low-frequency components of the multi-scale features based on incremental training (is considered to be a mathematical or mental step), and memorizing knowledge learned by the parallel convolutional neural networks in old tasks (is considered to be a mental process);
adaptively learning different fusion parameters according to different time spans of the input multi-scale data by using a multi-scale feature adaptive fusion technology, so as to enhance the ability to learn new tasks (is considered to be a mathematical step); and
outputting a Nino3.4 index reflecting a change rule of El Nino through fully connected layers according to the adaptively fused features, establishing a mapping function of an extreme rainfall probability r based on the Nino3.4 index, and in response to predicting that the value r goes beyond a threshold value k (is considered to be a mathematical step), carrying out rainstorm early warning (is considered to be a mental process), and
These mathematical and mental steps represent that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind.
Similar limitations comprise the abstract ideas of the independent claims 7.
Next, under the step 2A, prong two, it is considered whether the claim that recites a judicial exception is integrated into a practical application.
In this step, it is evaluated whether the claim recites meaningful additional elements that integrate the exception into a practical application of that exception.
In claim 1, the additional element in the preamble of “A weather warning method…” is not qualified for a meaningful limitation because it only generally links the use of the judicial exception to a particular technological environment or field of use. The additional elements/steps “carrying out rainstorm prevention and control of transmission lines in advance” as recited in the claims is considered to be an insignificant extra solution activity due to its unclear language or meaning as discussed above in the section clam rejection 35 U.S.C 112 (b).
In claim 7, the additional elements/steps recite the similar additional elements/steps as of claim 1. In claim 7, the additional element in the preamble of “A weather early device…” is not qualified for a meaningful limitation because it only generally links the use of the judicial exception to a particular technological environment or field of use. The additional elements/steps “carrying out rainstorm prevention and control of transmission lines in advance” as recited in the claims is considered to be an insignificant extra solution activity due to its unclear language or meaning as discussed above in the section claim rejection 35 U.S.C 112 (b).
In conclusion, the above additional elements, considered individually and in combination with the other claim elements do not reflect an improvement to other technology or technical field, and, therefore, do not integrate the judicial exception into a practical application. Therefore, the claims are directed to a judicial exception and require further analysis under the step 2B.
Considering the claim as a whole, one of ordinary skill in the art would not know the practical application of the present invention since the claims do not apply or use the judicial exception in some meaningful way.
The independent claims, therefore, are not patent eligible.
With regards to the dependent claims, the claims 2-6, 8-20 comprise the analogous subject matter and also comprise additional features/steps which are the part of an expanded abstract idea of the independent claim 1 and 7 (additionally comprising mathematical relationship/mental process steps) and, therefore, the dependent claims are not eligible without additional elements that reflect a practical application and qualified for significantly more for substantially similar reason as discussed with regards to independent claims.
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.
Claim(s) 1, 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tocornal et al. (US 20200348448 A1) herein after “Tocornal” in view of Rahaman, Baratin, Arpit, et al., published at ICML 2019 (PMLR, Vol. 97, pp. 5301-5310), herein after “Rahaman”
Regarding claim 1, Tocornal teaches a weather warning method based on incremental learning (para [0007] Embodiments of the present invention are in the field of climate forecasting, and pertain particularly to methods and systems for climate forecasting using an artificial neural network-based forecasting model.), comprising the steps:
down-sampling marine data to obtain multi-scale marine data (para [0053] FIG. 1 is a schematic diagram 100 illustrating grid cells used by an atmosphere-ocean coupled global climate model (AOGCM), and physical processes considered within each grid cell (From the National Oceanic & Atmospheric Administration, Geophysical Fluid Dynamics Laboratory). In this AOGCM, the earth 110 is divided into 3D grids 115 according to latitude, longitude, and height or pressure, and a pull-out image 120 shows different processes that may be modeled within each grid cell to calculate the evolution of the climate system, with interactions among neighboring cells imposed as boundary conditions. Pull-out image 120 illustrates various components that are taken into account by the AOGCM).
Examiner views the pull-out image 120 as a down sampling marine data out of 3D grids 115 to obtain a multi-scale data (i.e., see Fig. 1, 120 provides a multiscale marine data. Also see Fig. 5) and dividing the multi-scale data into a plurality of task sequences bounded by a preset year (para [0093] In this illustrative example, NN 500 may be setup to take, as input, a 24-month time series of monthly surface temperatures on a global 192×96 map or grid, and to forecast the Nino-3.4 index at a specified target lead-time such as 6 months into the future. FIG. 5B shows a graphical representation 550 of an exemplary data input, such as input climate data image 510, into NN 500, according to some embodiments of the present invention… This two-dimensional grid 550 in FIG. 5B is of size 26×14, much smaller than 192×96, for illustrative purpose only.);
Examiner views the input data (i.e., multi-scale data) is divided into 24-month time series of monthly temperature data for a neural network 500(i.e., plurality of task sequences bounded by a preset year, 24-month/2 year)
inputting the task sequences into parallel convolutional neural networks in a data flow form (para [0092] NNs can be viewed as parallel, densely interconnected computational models that adaptively learn through automatic adjustment of system parameters based on training data. Input information are modified based on system parameters when traversing through layers of interconnected neurons or nodes, to activate or trigger particular outputs.
[0093] FIG. 5B shows a graphical representation 550 of an exemplary data input, such as input climate data image 510, into NN 500, according to some embodiments of the present invention.
[0094] NN 500 is a Convolutional Recurrent Neural Network (CRNN) with a convolutional and recurrent architecture: it encodes the spatial information of each global surface temperature grid using a Convolutional Neural Network (CNN) first, then feeds the encoded information into a Recurrent Neural Network (RNN) having Long Short-Term Memory (LSTM) layers to learn from the temporal sequence),
Examiner views the NN 500 as parallel convolutional neural networks which takes the time series of data (i.e., task sequences in a data flow form)
and extracting multi-scale features through supervised representation learning (para [0095] More specifically, NN 500 first feeds 2-dimensional inputs 510 through multiple convolution (Conv2D) layers with Rectified Linear Units (ReLU), then a fully connected (FC) layer 520….Successive convolution-ReLU-pooling stages allow the successive extraction of low-level to high-level features, from local temperature correlations to distant teleconnections);
Herein examiner views the Convolution neural network as a supervised representation learning that extracts low to high level features of input data.
memorizing knowledge learned by the parallel convolutional neural networks in old tasks (para [0033] FIG. 5C is an exemplary Long Short-Term Memory (LSRM) cell for use in the CRNN in FIG. 5A;)
Examiner views the LSRM in prior art memorizes the knowledge learned by the parallel convolution neutral network from old task.
adaptively learning different fusion parameters according to different time spans of the input multi-scale data by using a multi-scale feature adaptive fusion technology, so as to enhance the ability to learn new tasks (para [0085] Such determinations may result in the construction of an optimization, convergence, forecast, or projection from a set of stored observational and/or simulation data. For example, components disclosed herein may employ various prediction and classification schemes and/or systems (e.g., support vector machines, neural networks, expert systems, Bayesian belief networks, fuzzy logic, data fusion engines, etc.) in connection with performing automatic and/or determined action in connection with the claimed subject matter. Thus, schemes and/or systems as disclosed herein may be used to automatically learn and perform a number of functions, actions, and/or determinations.
(para [0092] NNs can be viewed as parallel, densely interconnected computational models that adaptively learn through automatic adjustment of system parameters based on training data. Input information are modified based on system parameters when traversing through layers of interconnected neurons or nodes, to activate or trigger particular outputs.
Examiner views the data fusion engine with neural network provide adaptively learning of input data, source or nodes, layers (i.e., different fusion parameters) accordingly at different time spans of the input multi-scale data (since input data flows is sequential or in series, they have different time spans) by using a data fusion engine (i.e., multi-scale feature adaptive fusion technology), to automatically learn and perform functions or determinations (i.e., enhance the ability to learn new tasks) and
outputting a Nino3.4 index reflecting a change rule of El Nino through fully connected layers according to the adaptively fused features (para [0095] Output vector 525 from fully connect layer 520 may feed into an RNN 530 in sequences of successive months, such as 24 months. RNN 530 may have a many-to-one architecture, and may use two LSTM layers, each having 500 hidden units. At the end, the hidden state of the last time step may be decoded to a real value using another fully connected layer, to output predicted monthly Nino-3.4 sea surface temperature anomalies;)
Herein examiner views the Nino 3.4 index prediction and its calculation process(i.e., change rule of El nino) through connected layers is implemented according to a decoding technique (i.e., adaptively fused feature) , establishing a mapping function of an extreme rainfall probability r based on the Nino3.4 index (para [0105] Forecast targets 718 may include, but are not limited to, a scalar or vector target output climate variable to be predicted, a target lead-time at which the target output climate variable is to be predicted, and confidence levels for the forecasting results. Examples of a target output climate variable may include average monthly temperature and precipitation, average daily minimum temperature, seasonal minimum temperature, sea surface temperature, annual maximum wind speed, an index of a climate event such as El Nino Southern Oscillation (ENSO), and the like. A target lead-time may be on a monthly or a yearly scale, such as 3-months ahead, 6-months ahead, or 1-year ahead.
Para [0107] In some embodiments, a forecast skill score is computed for each validated GCM based on a forecast function, where the forecast function may be a data predictor function on the target output variable,
para [0108] The data predictor function on the target output variable may predict the target output variable such as the Nino-3.4 index directly from the GCM data, or may predict another closely related climate variable, such as the Nino-3 index and/or the Nino-4 index.
Para [0126] These indices measure the state of the warm and cold temperature cycles in the Equatorial Pacific Ocean, the Atlantic Ocean, and the Northern Pacific Ocean, and affect climate variables such as temperature and precipitation throughout the world.)
From above examiner views the predictor function (i.e., establishing a mapping function) for target output (i.e., precipitation or an extreme rainfall probability r) based on the Nino3.4 indexes , and in response to predicting that the value r goes beyond a threshold value k (para [0111] Forecast skill may be measured by a mean square error (MSE), a correlation between the forecast and the actual values of the target climate variable, or other appropriate error or distance metrics. Such MSE or correlation value computed for the forecast function discussed above may be viewed as a forecast skill score… Again, forecast skill scores may be compared to a threshold;), carrying out rainstorm early warning (para [0105] Examples of a target output climate variable may include average monthly temperature and precipitation, average daily minimum temperature, seasonal minimum temperature, sea surface temperature, annual maximum wind speed, an index of a climate event such as El Nino Southern Oscillation (ENSO), and the like. A target lead-time may be on a monthly or a yearly scale, such as 3-months ahead, 6-months ahead, or 1-year ahead.) and carrying out rainstorm prevention and control of transmission lines in advance (para [0045] FIG. 15 is an illustrative diagram showing a United States map of hydroelectric plants, and seasonal risk predictions at a selected hydroelectric plant, according to some embodiments of the present invention; and).
Examiner views weather risk prediction provide warning or indication of severe weather (i.e., rainfall) provide preventative measures to be taken at hydroelectric plants (i.e., control of transmission lines in advance.)
Tocornal does not clearly teach selectively constraining, by a multi-scale feature frequency domain distillation technology, drift of low-frequency components of the multi-scale features based on incremental training.
Rahman teaches selectively constraining, by a multi-scale feature frequency domain distillation technology, drift of low-frequency components of the multi-scale features based on incremental training (page 4, Fig. 1 Left (a,b)…Gist: We find that even when higher frequencies have larger amplitudes, the model prioritizes learning lower frequencies first. We also find that the spectral norm of weights increases as the model fits higher frequency, which is what we expect from Theorem 1.)
The above paragraph does not explicitly say a frequency domain distillation technology; however, examiner views the above paragraph also provides an idea of frequency domain distillation technology as claimed- prioritizing lower frequencies first (i.e., selectively constraining drift or loss of low frequency components) in a spectrum feature (includes both low and high frequency i.e., multi-scale feature) where training is incremental in neural network.
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have incorporated the idea of Rahman into Tocornal for the purpose of selectively learning the lower frequency component during an incremental training of neural network so that a stronger memory and deeper understanding of a studied feature can be obtained.
Claim 7 is rejected as claim 1 having same claim limitation.
Claim(s) 3, 12, 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over the combination of Tocornal and Rahaman in view of Dai, Yimian, et al. "Attentional Feature Fusion." 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), IEEE, 2021, pp. 3559-68. IEEE Xplore, doi:10.1109/WACV48608.2021.00359.
Regarding claim 3, The combination of Tocornal and Rahaman teach the weather warning method based on incremental learning according to claim 1, Tocornal teaches wherein the multi-scale feature adaptive fusion technology comprises: multi-scale parallel networks ( para [0092] NNs can be viewed as parallel, densely interconnected computational models that adaptively learn through automatic adjustment of system parameters based on training data.), and one adaptive fusion function (para [0085] For example, components disclosed herein may employ various prediction and classification schemes and/or systems (e.g., support vector machines, neural networks, expert systems, Bayesian belief networks, fuzzy logic, data fusion engines, etc.) in connection with performing automatic and/or determined action in connection with the claimed subject matter).
Tocornal does not clearly teach two bottleneck layers, two fully connected layers and one adaptive fusion function.
Dai, Yimian teaches bottleneck layers, two fully connected layers and one adaptive fusion function.
bottleneck layer, two fully connected layers (page 3561, left col. Section 3.1. This is achieved by a bottleneck with two fully connected (FC) layers).
Dai, Yimian teaches a bottleneck layer. Examiner views the person skilled in the art would modify neural network with two bottleneck layers with two fully connected layers to improve data compression, neural network learning features and maintain the efficiency of neural network.
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing of the invention to have modified and incorporated the idea of Dai, Yimian into Tocornal for the purpose of having two bottlenecks with two fully connected layers so an improved data compression, an improved neural network learning feature and an improved efficiency of neural network can be obtained leading to an accurate climate or weather forecast.
Claim 12 and 17 is rejected as claim 3 having same claim limitation.
Allowable Subject Matter
There are no prior art rejections for claims 2, 4-6, 8-11, 14-16 and 18-20. However, examiner cannot comment on their allowability until the rejections under 35 USC 112 and 101 are adequately addressed.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Suresh et al (US 10895802 B1) discusses deep learning for weather forecast in port operations.
Tsui et al (US 20030105597 A1) discuss generating sequences of data in machine learning.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHARAD TIMILSINA whose telephone number is (571)272-7104. The examiner can normally be reached Monday-Friday 9:00-5:00.
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, Catherine Rastovski can be reached at 571-270-0349. 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.
/SHARAD TIMILSINA/Examiner, Art Unit 2863
/Catherine T. Rastovski/Supervisory Primary Examiner, Art Unit 2863