Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are: "an information determination module", "a probability model determination module", "a weight determination module", "a model establishment module", and "a prediction module" in claim 13.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 14 rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
As to claim 14,
The claim does not limit the recited “readable medium” to a non-transitory computer-readable storage medium. Under the broadest reasonable interpretation, the claimed “readable medium” encompasses transitory forms of signal transmission, such as a propagating electrical or electromagnetic signal or carrier wave. This rejection may be overcome by amending claim 14 to recite, for example, “a non-transitory computer-readable storage medium” instead of “a readable medium,” provided there is proper support in the specification.
Claim 1-4, and 13-15 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
MPEP 2106 (III) sets out steps for evaluating whether a claim is drawn to patent-eligible subject matter. The analysis of claims 1-20, in accordance with these steps, follows.
Step 1 Analysis:
Claims 1-4 are directed to method (processes). Claims 13 and 15 are directed to a computer apparatus (machine). Therefore, claims 1-4, and 13 and 15 fall into one of four statutory categories (i.e., process, machine, article of manufacture).
Claim 14 is not directed to a statutory category (see U. S. C. 101 rejection for claim 14 above because the claim is directed to non-statutory subject matter.)
As to claim 1,
Step 2A Prong 1: this claim recites the following abstract ideas:
determining feature information of a target device and detection point data information corresponding to the target device; (the limitation describes identifying and evaluating device characteristics and data information, which is an evaluation and judgment activity that can be performed as a mental process in the human mind)
determining, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of feature data of the target device and a probability distribution model of detection point data with unshared data; (the limitation describes computing probability distributions from data, which is a mental process implemented using a pen and paper.)
determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data; (the limitation describes calculating a weight value based on probability distributions, which is a mental process implemented using a pen and paper.)
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
establishing a federated learning model according to the unshared data, the weight of the unshared data and a device fault label corresponding to the unshared data; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
predicting a device fault of the target device according to the federated learning model; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
The additional element does not integrate the judicial exception into practical application and does not amount to significantly more than the Judicial exception.
As to claim 2,
Step 2A Prong 1: this claim recites the following abstract ideas:
determining a target device distribution probability of the unshared data according to the probability distribution model of the feature data; (the limitation describes determining a probability value from a distribution model, which is a mental process implemented using a pen and paper)
determining a detection point distribution probability of the unshared data according to the probability distribution model of the detection point data; (the limitation describes determining a probability value from a distribution model, which is a mental process implemented using a pen and paper)
determining a ratio of the target device distribution probability of the unshared data to the detection point distribution probability of the unshared data as the weight of the unshared data; (the limitation describes determining a ratio of two probability values, which is a mental process implemented using a pen and paper)
Step 2A Prong 2 and 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim is ineligible.
As to claim 3,
Step 2A Prong 1: this claim recites the following abstract ideas:
determining, based on the feature information of the target device and the detection point data information corresponding to the target device, the feature data of the target device and the detection point data with the unshared data; (the limitation describes identifying and categorizing data, which is an evaluation and judgment activity that can be performed as a mental process in the human mind)
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
calculating data distribution of the feature data according to a feature data parameter model, and determining the feature data parameter model with determined parameters as the probability distribution model of the feature data; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
calculating data distribution of the detection point data according to a detection point data parameter model, and determining the detection point data parameter model with determined parameters as the probability distribution model of the detection point data; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
The additional element does not integrate the judicial exception into practical application and does not amount to significantly more than the Judicial exception.
As to claim 4,
Step 2A Prong 1: This claim does not recite an additional abstract idea, but the claim depends on claim 1,
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
determining a detection point local model according to the unshared data, the weight of the unshared data and the device fault label corresponding to the unshared data; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
establishing the federated learning model according to at least two detection point local models; (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
The additional element does not integrate the judicial exception into practical application and does not amount to significantly more than the Judicial exception.
As to claim 13,
Step 2A Prong 1: this claim recites the following abstract ideas:
determine feature information of a target device and detection point data information corresponding to the target device; (the limitation describes identifying and evaluating device characteristics and data information, which is an evaluation and judgment activity that can be performed as a mental process in the human mind)
determine, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of feature data of the target device and a probability distribution model of detection point data with unshared data; (the limitation describes computing probability distributions from data, which is a mental process implemented using a pen and paper.)
determine a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data; (the limitation describes calculating a weight value based on probability distributions, which is a mental process implemented using a pen and paper.)
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
an information determination module configured (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
a probability model determination module configured (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
a weight determination module configured (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
a model establishment module configured to establish a federated learning model according to the unshared data, the weight of the unshared data and a device fault label corresponding to the unshared data; and (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
a prediction module configured to predict a device fault of the target device according to the federated learning model. (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
The additional element does not integrate the judicial exception into practical application and does not amount to significantly more than the Judicial exception.
As to claim 14,
Step 2A Prong 1: This claim does not recite an additional abstract idea, but the claim depends on claim 1
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
A readable medium, comprising execution instructions, when a processor of an electronic device executes the execution instructions, the electronic device (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
As to claim 15,
Step 2A Prong 1: This claim does not recite an additional abstract idea, but the claim depends on claim 1
Step 2A Prong 2 and 2B: the claim recited the following additional elements:
An electronic device, comprising a processor and a memory storing execution instructions, when the processor executes the execution instructions stored by the memory, the processor performing the method (This limitation is directed to mere instruction to apply the abstract idea on a generic computer to process, which is a well-understood, routine, conventional activity, see MPEP 2106.05(d)(II)(i))
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) 1-4, and 13-15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Hu et al. (US 20230106985 A1) in view of Vasudevan et al. (US 20200104710 A1) and Inagaki et al. (US 20170031329 A1).
As to claim 1, Hu teaches:
determining, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of feature data of the target device and a probability distribution model of detection point data with unshared data. (See Hu paragraph [0048] "the methods 100, 200 require the transmission from local nodes of a representation of the distribution of data within their local data sets. This maintains the privacy advantages of conventional federated learning, as the data itself is not transmitted", and see Hu paragraph [0108] "Algorithm 1 … Input: Dataset D.sub.k = {x.sub.1, x.sub.2,...,x.sub.mk} where m.sub.k is number of samples for client k, D.sub.0 is global dataset where others are distributed datasets ... 4: Train a G.sub.k(x) = Σ.sub.i=1.sup.gk α.sub.iN(x|μ.sub.ki, Σ.sub.k,) to approximate D.sub.k, where g.sub.k is the number of Gaussian components”)
Examiner note: A Gaussian mixture probability distribution model is trained for the global/reference dataset D₀ (target device feature data) and for each distributed local dataset D₁… D_K (detection point data), where the local data itself is never transmitted i.e., unshared data.
establishing a federated learning model according to the unshared data. (See Hu paragraph [0041] "the method comprises, for at least one learning group, at each of the plurality of distributed nodes within said learning group, developing a node version of the machine-learning model, based on the seed version of the machine-learning model and the associated local data set, and using the machine-learning algorithm.”)
Hu does not explicitly teach:
determining feature information of a target device and detection point data information corresponding to the target device;
a device fault label corresponding to the unshared data; and
predicting a device fault of the target device according to the federated learning model.
However, Inagaki teaches:
determining feature information of a target device and detection point data information corresponding to the target device. (See Inagaki paragraph [0041] "The fault prediction system 1 further includes a sensor 11 which detects the state of the robot 2 or the surrounding environment. The sensor 11 may include at least one of a force sensor, a torque sensor, a vibration sensor, a sound collection sensor, an image sensor, a distance sensor, a temperature sensor, a humidity sensor, a flow sensor, a light quantity sensor, a pH sensor, a pressure sensor, a viscosity sensor, and an odor sensor. Data output from the sensor 11 (to be also simply referred to as 'output data' hereinafter) is input to a state observation unit 52 of the machine learning device 5.”)
Examiner note: The sensors detecting the state of the industrial machine are the detection points, and the observed sensor outputs constitute the feature information of the target device.
a device fault label corresponding to the unshared data. (See Inagaki paragraph [0045] "The determination data is defined as data used to determine whether a fault has occurred or the degree of fault.”, and see Inagaki paragraph [0053] "The neural network learns the relationship between the state variable observed by the state observation unit 52 and the occurrence of a fault, i.e., fault conditions, by so-called supervised learning, in accordance with a training data set generated based on a combination of this state variable and the determination data obtained by the determination data obtaining unit 51. In supervised learning, a large number of sets of data of certain inputs and results (labels) are fed into a machine learning device”)
Examiner note: The fault determination data is expressly the label paired with each training sample in the training data set; in the combination, this fault-labeled sensor data is the local data set held (unshared) at Hu's distributed nodes.
predicting a device fault of the target device according to the federated learning model. (See Inagaki paragraph [0065] "The fault information output unit 42 outputs the fault information of the robot 2 in response to input of the state variable via the state observation unit 41, based on the result of learning by a learning unit 53 of the above-mentioned machine learning device 5 in accordance with a training data set.”)
examiner note: The fault prediction is performed according to the learned model, which Inagaki teaches may be the same model shared across plural machines.
Hu and Inagaki do not explicitly teach:
determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data;
the weight of the unshared data.
However, Vasudevan teaches:
determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data. (See Vasudevan paragraph [0033] "Intuitively, P.sub.t(y) describes the distribution of outputs in the target dataset, and P.sub.t(y)/P.sub.s(y) reweights object classes during the pre-training of the source neural network 120 so that the class distribution statistics match P.sub.y(t). P.sub.t(y)/P.sub.s(y) is referred to as an importance weight associated with a source output y.”)
Examiner note: The importance weight is determined according to the target-side distribution and the source-side (detection point) distribution; in the combination, the distribution inputs are the GMM models Hu computes for the reference dataset and each local dataset.
the weight of the unshared data. (See Vasudevan paragraph [0038] "The system 100 generates the pre-training dataset 106 by sampling a set of source training examples from the source dataset 104 based on the computed importance weights for the source outputs.”)
Examiner note: The model is trained according to the computed importance weights of the source (unshared) training data.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Hu to include determining feature information of a target device and detection point data information via sensors, a device fault label corresponding to the unshared data, and predicting a device fault of the target device according to the learning model, as taught by Inagaki, in order to achieve "accurate fault prediction even when factors which may lead to faults are complicated and make it difficult to preset fault conditions" (Inagaki paragraph [0060]) and to learn "accurate fault conditions according to the actual circumstances of use" (Inagaki paragraph [0073]).
Further, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to further modify the combination of Hu and Inagaki to include determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data, and establishing the learning model according to that weight, as taught by Vasudevan, in order to reweight the unshared training data "so that the class distribution statistics match" the target distribution (Vasudevan paragraph [0033]), thereby making training over dissimilar distributed datasets behave as training matched to the target device's data.
Therefore, claim 1 would have been obvious over Hu in view of Inagaki and further in view of Vasudevan.
As to claim 2. Hu-Inagaki as modified by Vasudevan teaches the method according to claim 1,
wherein the step of determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data comprises: (see Vasudevan paragraph [0033] "Intuitively, P.sub.t(y) describes the distribution of outputs in the target dataset, and P.sub.t(y)/P.sub.s(y) reweights object classes during the pre-training of the source neural network 120 so that the class distribution statistics match P.sub.y(t). P.sub.t(y)/P.sub.s(y) is referred to as an importance weight associated with a source output y.”)
determining a target device distribution probability of the unshared data according to the probability distribution model of the feature data; (see Vasudevan paragraph [0036] "For each source output y in the set of possible source outputs, the system 100 determines P.sub.t(y) that represents a rate of appearance of the source output y in the set of temporary predicted outputs that have been generated by the trained classifier neural network 130 for the target training inputs in the target dataset 102.”)
determining a detection point distribution probability of the unshared data according to the probability distribution model of the detection point data; and (see Vasudevan paragraph [0034] "To determine P.sub.s(y), the system 100 determines a rate of appearance of the source output y in the source dataset D.sub.s by dividing the number of times the source output y appears by the total number of source training examples in the source dataset D.sub.s (104).”)
determining a ratio of the target device distribution probability of the unshared data to the detection point distribution probability of the unshared data as the weight of the unshared data. (Vasudevan paragraph [0037]: "After computing P.sub.s(y) and P.sub.t(y) for each of the source outputs in the source dataset 104, the system 100 computes the importance weight P.sub.t(y)/P.sub.s(y) for each source output.”)
As to claim 3, Hu-Inagaki as modified by Vasudevan teaches the method according to claim 1,
wherein the step of determining, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of feature data of the target device and a probability distribution model of detection point data with unshared data comprises: (see Hu paragraph [0048] "the methods 100, 200 require the transmission from local nodes of a representation of the distribution of data within their local data sets. This maintains the privacy advantages of conventional federated learning, as the data itself is not transmitted", and see Hu paragraph [0108] "Algorithm 1 … Input: Dataset D.sub.k = {x.sub.1, x.sub.2,...,x.sub.mk} where m.sub.k is number of samples for client k, D.sub.0 is global dataset where others are distributed datasets ... 4: Train a G.sub.k(x) = Σ.sub.i=1.sup.gk α.sub.iN(x|μ.sub.ki, Σ.sub.k,) to approximate D.sub.k, where g.sub.k is the number of Gaussian components”)
determining, based on the feature information of the target device and the detection point data information corresponding to the target device, the feature data of the target device and the detection point data with the unshared data; (see Hu paragraph [0108] "Algorithm 1 … Input: Dataset D.sub.k = {x.sub.1, x.sub.2,...,x.sub.mk} where m.sub.k is number of samples for client k, D.sub.0 is global dataset where others are distributed datasets 2: for client X = 0,1,2,...K do Statistic gathering for K+1 clients”)
Examiner note: The reference dataset D₀ corresponds to the target device feature data and the distributed client datasets to the unshared detection point data; both are determined as inputs to the statistic-gathering step.
calculating data distribution of the feature data according to a feature data parameter model, and determining the feature data parameter model with determined parameters as the probability distribution model of the feature data; and (see Hu paragraph [0108] “Algorithm 1… 2: for client X = 0,1,2,...K do Statistic gathering for K+1 clients 3: Gather quantity of labels q.sub.k per category 4: Train a G.sub.k(x) = Σ.sub.i=1.sup.gk α.sub.iN(x|μ.sub.ki, Σ.sub.k,) to approximate D.sub.k, where g.sub.k is the number of Gaussian components”)
Examiner note: Because the loop begins at client 0, the Gaussian mixture parameter model G₀(x) is trained on the global/reference dataset D₀ (the target device feature data); the GMM with its determined parameters (α_i, μ_ki, Σ_k) is the probability distribution model of the feature data.
calculating data distribution of the detection point data according to a detection point data parameter model, and determining the detection point data parameter model with determined parameters as the probability distribution model of the detection point data. (See Hu paragraph [0108] “Algorithm 1 … 4: Train a G.sub.k(x) = Σ.sub.i=1.sup.gk α.sub.iN(x|μ.sub.ki, Σ.sub.k,) to approximate D.sub.k, where g.sub.k is the number of Gaussian components 5: Upload dataset representation C.sub.k = [q.sub.k, G.sub.k(x)) to the global server"
Examiner note: Each detection point (distributed client) dataset D_k is fitted with a Gaussian mixture parameter model G_k(x) whose determined parameters constitute the probability distribution model of that detection point's data.
As to claim 4, Hu-Inagaki as modified by Vasudevan teaches the method according to claim 1,
wherein the step of establishing a federated learning model according to the unshared data, (see Hu paragraph [0041] "the method comprises, for at least one learning group, at each of the plurality of distributed nodes within said learning group, developing a node version of the machine-learning model, based on the seed version of the machine-learning model and the associated local data set, and using the machine-learning algorithm.”)
determining a detection point local model according to the unshared data, the weight of the unshared data and the device fault label corresponding to the unshared data; and (see Hu paragraph [0041] "developing a node version of the machine-learning model, based on the seed version of the machine-learning model and the associated local data set, and using the machine-learning algorithm. The node version of the machine-learning model is a version of the model that is unique to the node, having been developed by the node starting from the seed version of the model and using the machine-learning algorithm and the local data set associated with the node.”)
establishing the federated learning model according to at least two detection point local models. (see Hu paragraph [0041] "the method comprises, for at least one learning group, at each of the plurality of distributed nodes within said learning group, developing a node version of the machine-learning model, based on the seed version of the machine-learning model and the associated local data set, and using the machine-learning algorithm.”, and see Hu paragraph [0099]: "The Master Node then combines the obtained node versions of the machine-learning model to form a group version of the machine learning model for the learning group in step 926. For example, the Master node may average each of the obtained node versions of the machine-learning model to form the group version of the machine-learning model.”)
Hu do not explicitly teach:
the weight of the unshared data and
a device fault label corresponding to the unshared data comprises:
However, Vasudevan teaches:
the weight of the unshared data and (see Vasudevan paragraph [0038] "The system 100 generates the pre-training dataset 106 by sampling a set of source training examples from the source dataset 104 based on the computed importance weights for the source outputs.”)
Hu and Vasudevan do not explicitly teach:
a device fault label corresponding to the unshared data comprises
However, Inagaki teaches
a device fault label corresponding to the unshared data comprises: (see Inagaki paragraph [0045] "The determination data is defined as data used to determine whether a fault has occurred or the degree of fault.”, and see Inagaki paragraph [0053] "The neural network learns the relationship between the state variable observed by the state observation unit 52 and the occurrence of a fault, i.e., fault conditions, by so-called supervised learning, in accordance with a training data set generated based on a combination of this state variable and the determination data obtained by the determination data obtaining unit 51. In supervised learning, a large number of sets of data of certain inputs and results (labels) are fed into a machine learning device”)
As to claim 13, Hu teaches
a probability model determination module configured to determine, based on the feature information of the target device and the detection point data information corresponding to the target device, a probability distribution model of feature data of the target device and a probability distribution model of detection point data with unshared data; (see Hu paragraph [0048] "the methods 100, 200 require the transmission from local nodes of a representation of the distribution of data within their local data sets. This maintains the privacy advantages of conventional federated learning, as the data itself is not transmitted", and see Hu paragraph [0108] "Algorithm 1 … Input: Dataset D.sub.k = {x.sub.1, x.sub.2,...,x.sub.mk} where m.sub.k is number of samples for client k, D.sub.0 is global dataset where others are distributed datasets ... 4: Train a G.sub.k(x) = Σ.sub.i=1.sup.gk α.sub.iN(x|μ.sub.ki, Σ.sub.k,) to approximate D.sub.k, where g.sub.k is the number of Gaussian components”)
a model establishment module configured to establish a federated learning model according to the unshared data, (see Hu paragraph [0041] "the method comprises, for at least one learning group, at each of the plurality of distributed nodes within said learning group, developing a node version of the machine-learning model, based on the seed version of the machine-learning model and the associated local data set, and using the machine-learning algorithm.”)
Hu does not explicitly teach:
an information determination module configured to determine feature information of a target device and detection point data information corresponding to the target device;
a device fault label corresponding to the unshared data; and
a prediction module configured to predict a device fault of the target device according to the federated learning model.
However, Inagaki teaches:
an information determination module configured to determine feature information of a target device and detection point data information corresponding to the target device; (see Inagaki paragraph [0041] "The fault prediction system 1 further includes a sensor 11 which detects the state of the robot 2 or the surrounding environment. The sensor 11 may include at least one of a force sensor, a torque sensor, a vibration sensor, a sound collection sensor, an image sensor, a distance sensor, a temperature sensor, a humidity sensor, a flow sensor, a light quantity sensor, a pH sensor, a pressure sensor, a viscosity sensor, and an odor sensor. Data output from the sensor 11 (to be also simply referred to as 'output data' hereinafter) is input to a state observation unit 52 of the machine learning device 5.”)
a device fault label corresponding to the unshared data; and (see Inagaki paragraph [0045] "The determination data is defined as data used to determine whether a fault has occurred or the degree of fault.”, and see Inagaki paragraph [0053] "The neural network learns the relationship between the state variable observed by the state observation unit 52 and the occurrence of a fault, i.e., fault conditions, by so-called supervised learning, in accordance with a training data set generated based on a combination of this state variable and the determination data obtained by the determination data obtaining unit 51. In supervised learning, a large number of sets of data of certain inputs and results (labels) are fed into a machine learning device”)
a prediction module configured to predict a device fault of the target device according to the federated learning model. (See Inagaki paragraph [0065] "The fault information output unit 42 outputs the fault information of the robot 2 in response to input of the state variable via the state observation unit 41, based on the result of learning by a learning unit 53 of the above-mentioned machine learning device 5 in accordance with a training data set.”)
Hu and Inagaki do not explicitly teach:
a weight determination module configured to determine a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data;
the weight of the unshared data.
However, Vasudevan teaches:
a weight determination module configured to determine a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data; (see Vasudevan paragraph [0033] "Intuitively, P.sub.t(y) describes the distribution of outputs in the target dataset, and P.sub.t(y)/P.sub.s(y) reweights object classes during the pre-training of the source neural network 120 so that the class distribution statistics match P.sub.y(t). P.sub.t(y)/P.sub.s(y) is referred to as an importance weight associated with a source output y.”)
the weight of the unshared data and (see Vasudevan paragraph [0038] "The system 100 generates the pre-training dataset 106 by sampling a set of source training examples from the source dataset 104 based on the computed importance weights for the source outputs.”)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Hu to include determining feature information of a target device and detection point data information via sensors, a device fault label corresponding to the unshared data, and predicting a device fault of the target device according to the learning model, as taught by Inagaki, in order to achieve "accurate fault prediction even when factors which may lead to faults are complicated and make it difficult to preset fault conditions" (Inagaki paragraph [0060]) and to learn "accurate fault conditions according to the actual circumstances of use" (Inagaki paragraph [0073]).
Further, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to further modify the combination of Hu and Inagaki to include determining a weight of the unshared data according to the probability distribution model of the feature data and the probability distribution model of the detection point data, and establishing the learning model according to that weight, as taught by Vasudevan, in order to reweight the unshared training data "so that the class distribution statistics match" the target distribution (Vasudevan paragraph [0033]), thereby making training over dissimilar distributed datasets behave as training matched to the target device's data.
Therefore, claim 13 would have been obvious over Hu in view of Inagaki and further in view of Vasudevan.
As to claim 14, is directed to a computer program embodiment that corresponds to the method claim 1, see the rejection for claim 1 above, which also applies to claim 14. In addition, claim 14 recites the additional elements “A readable medium, comprising execution instructions, when a processor of an electronic device executes the execution instructions, the electronic device” (see Hu paragraph [0144] “The methods may also be carried out according to the instructions of a computer program, and the present disclosure also provides a computer readable medium having stored thereon a program for carrying out any of the methods described herein.”)
As to claim 15, is directed to a device embodiment that corresponds to the method claim 1, see the rejection for claim 1 above, which also applies to claim 15. In addition, claim 15 recites the additional elements “An electronic device, comprising a processor and a memory storing execution instructions, when the processor executes the execution instructions stored by the memory, the processor performing the method” (see Hu paragraph [0134] “comprises a processor or processing circuitry 1102, and may comprise a memory 1104 and interfaces 1106. The processing circuitry 1102 is operable to perform some or all of the steps of the method”
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ABDULLAH K ABOUD whose telephone number is (571)272-0025. The examiner can normally be reached Mon-Fri 8am-5pm.
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, Li B Zhen, can be reached at (571) 272-3768. 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.
/ABDULLAH KHALED ABOUD/ Examiner, Art Unit 2121
/Li B. Zhen/ Supervisory Patent Examiner, Art Unit 2121
Prosecution Insights