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 .
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(s) 1 – 15 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 pre-AIA the applicant regards as the invention.
Claim 1 recites:
“responsive to a valid trust signal from a first local device that is distinct from the apparatus, …”, there is insufficient antecedent basis for the limitation in the claim, which arises from the ambiguity of reference. Specifically, the term "… the apparatus" lack clarity because it is not clearly defined or linked to a prior element in the claim. For a more precise understanding, the claim should explicitly define or identify terms before refereeing it.
wherein the local version of the global model is generated using the training samples; there is insufficient antecedent basis for the limitation in the claim, which arises from the ambiguity of reference. Specifically, the term "… the training samples " lack clarity because it is not clearly defined or linked to a prior element in the claim. For a more precise understanding, the claim should explicitly define or identify terms before refereeing it.
… the local loss value being indicative of the local loss value being more preferred than the global loss value, renders the claim indefinite because the term "… more preferred " does not specify what criterion makes the local loss value more preferred than the global loss value. It is therefore unclear whether “more preferred” means lower loss, higher accuracy or another metric. Appropriate clarification is required.
The dependent claims inherit the same deficiency as the independent claims because they depend therefrom and do not add limitations sufficient to overcome the identified deficiency.
Claim 7 recites: … the local loss value being indicative of the local loss value being more preferred than the global loss value, renders the claim indefinite because the term "… more preferred " does not specify what criterion makes the local loss value more preferred than the global loss value. It is therefore unclear whether “more preferred” means lower loss, higher accuracy or another metric. Appropriate clarification is required.
The dependent claims inherit the same deficiency as the independent claims because they depend therefrom and do not add limitations sufficient to overcome the identified deficiency.
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 therefore, subject to the conditions and requirements of this title.
Claim(s) 1 – 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without significantly more.
In step 1, of the 101-analysis set forth in the MPEP 2106, the examiner has determined
that the following limitations recite a process that, under the broadest reasonable interpretation, falls within one or more statutory categories (processes).
In step 2A prong 1, of the 101-analysis set forth in MPEP 2106, the examiner has determined
that the following limitations recite a process that, under broadest reasonable interpretation, recites abstract idea but for the recitation of generic computer components:
Regarding claim 1,
analyzing the local loss value based on the quantity of training samples and the quantity of test samples; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves analyzing numerical information and determining a loss value based on quantities of samples). See (MPEP 2106.04).
responsive to a result of the analysis of the local loss value being indicative of the local loss value being more preferred than the global loss value:
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing local and global loss to check which one is preferable). See (MPEP 2106.04).
If the claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process, but for the recitation of generic computer components, then it falls within the mental process. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 of the 101-analysis, set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
responsive to a valid trust signal from a first local device that is distinct from the apparatus, communicating a global model and a global loss value to the first local device;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
receiving a local loss value, an indication of a quantity of training samples, and an indication of a quantity of test samples from the first local device, wherein the local loss value is based on execution of a local version of the global model, generated by the first local device, on a local test dataset by the first local device, wherein the local version of the global model is generated using the training samples;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
… communicating a trust signal to the first local device;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
responsive to communication of the trust signal to the first local device, receiving the local version of the global model from the first local device;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
updating the global model based on the local version of the global model, the local loss value, the quantity of training samples, and the quantity of test samples; and
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations which do not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
communicating the updated global model to the first local device and to a second local device that is distinct from the apparatus.
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
In Step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the
claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitation (V), recites mere application of the abstract idea or mere instructions to implement an abstract idea on a computer are deemed insufficient to transform the judicial exception to a patentable invention because the limitations generally apply the use of a generic computer and/or process with the judicial exception, see MPEP 2106.05(f).
Regarding limitations (I, II, III, IV and VI), additional elements considered extra/post solution activity, as analyzed above, are activities that are well-understood routine and conventional, specifically: the courts have recognized the computer functions as well‐understood, routine, and conventional functions.
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
As analyzed above, the additional elements, analyzed above, do not integrate the noted judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore, the claim is directed to an abstract idea.
Regarding claim 2, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
responsive to the local loss value from the first local device being more preferred than the global loss value, communicating the local loss value to the first local device and to the second local device as an updated global loss value associated with the updated global model.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyzed failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 3, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the local loss value from the first local device is a first local loss value, and wherein the method further comprises, in association with the analysis of the first local loss value
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
… determining whether the first local loss value is less than a second loss value from the second local device.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing two numerical values to determine which value is smaller). See (MPEP 2106.04).
Regarding claim 4, dependent upon claim 3, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
further comprising, responsive to determining that a third local loss value from a third local device that is distinct from the apparatus is less than the global loss value:
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing a local loss value with a global loss value to determine whether a specified numerical condition is satisfied.) See (MPEP 2106.04).
updating the global model based on a different local version of the global model from the third local device and the third local loss value;
Deemed insufficient to transform the judicial exception to a patentable invention because the limitation is directed to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea and are considered to adding the words “apply it” (or an equivalent) with the judicial exception, See MPEP 2106.05(f).
Limitations directed to using the computer as a tool for implementing an abstract idea cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
communicating the updated global model to the third local device.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 5, dependent upon claim 4, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
communicating the trust signal to the third local device; and responsive to communication of the trust signal to the third local device, receiving the different local version of the global model from the third local device.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 6, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the method is performed by a cloud server.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Regarding claim 7, In step 2A prong 1,
determine whether the first local loss value or the second local loss value is more preferred than a global loss value associated with the global model;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing local loss values with a global loss value to determine which values are more favorable). See (MPEP 2106.04).
If the claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process, but for the recitation of generic computer components, then it falls within the mental process. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 of the 101-analysis, set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
A host system for federated learning, configured to: communicate a global model to a first local device and a second local device, wherein the first local device and the second local device are distinct from and trusted by the host system; receive from the first local device, a first local loss value based on execution of a first local version of the global model on a first quantity of samples by the first local device; receive from a second local device, a second local loss value based on execution of a second local version of the global model on a second quantity of samples by the second local device;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
responsive to determining that the first local loss value is more preferred than the global loss value, communicate to the first local device and the second local device:
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
a first updated version of the global model based on the first local version of the global model; and a first updated global loss value based on the first local loss value; and
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations which do not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
… communicate to the first local device and the second local device:
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
a second updated version of the global model based on the second local version of the global model; and a second updated global loss value based on the second local loss value.
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations which do not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
In Step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the
claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitations (III and V), recite mere application of the abstract idea or mere instructions to implement an abstract idea on a computer are deemed insufficient to transform the judicial exception to a patentable invention because the limitations generally apply the use of a generic computer and/or process with the judicial exception, see MPEP 2106.05(f).
Regarding limitations (I, II and IV), additional elements considered extra/post solution activity, as analyzed above, are activities that are well-understood routine and conventional, specifically: the courts have recognized the computer functions as well‐understood, routine, and conventional functions.
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
As analyzed above, the additional elements, analyzed above, do not integrate the noted judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore, the claim is directed to an abstract idea.
Regarding claim 8, dependent upon claim 7, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
to communicate to the first local device and the second local device: a third updated version of the global model based on the first local version and the second local version of the global model; and a third updated global loss value based on the first local loss value and the second local loss value.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 9, dependent upon claim 7, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
determine whether a respective trust signal from the first local device and the second local device is valid;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves evaluating trust signals against a criterion to determine whether the signals are valid). See (MPEP 2106.04).
responsive to determining the respective trust signal from the first local device is valid, determine that the first local device is trusted; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves determining a trust status based on whether an associating trust signal satisfies a validity condition). See (MPEP 2106.04).
responsive to determining the respective trust signal from the second local device is valid, determine that the second local device is trusted.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves determining a trust status based on whether an associated trust signal satisfies a validity condition). See (MPEP 2106.04).
Regarding claim 10, dependent upon claim 9, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
determine whether the first local loss value is more preferred than the global loss value in response to determining that the first local device is trusted; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing a local loss value with a global loss value after evaluating the trust status of the local device). See (MPEP 2106.04).
determine whether the second local loss value is more preferred than the global loss value in response to determining that the second local device is trusted.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing a local loss value with a global loss value after evaluating the trust status of the local device). See (MPEP 2106.04).
Regarding claim 11, dependent upon claim 7, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
determine whether the first local loss value is more preferred than the global loss value based on the first quantity of samples; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves evaluating a local loss value relative to a global loss value based on sample quantities). See (MPEP 2106.04).
determine whether the second local loss value is more preferred than the global loss value based on the second quantity of samples.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves evaluating a local loss value relative to a global loss value based on sample quantities). See (MPEP 2106.04).
Regarding claim 12, dependent upon claim 7, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
responsive to determining that the first local loss value is more preferred than the global loss value, to communicate a trust signal from the host system to the first local device indicative of the host system requesting the first local version of the global model.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyzed failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 13, dependent upon claim 12, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
associate a time stamp with communicating the trust signal to the first local device; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves associating a temporal identifier with an event or communication). See (MPEP 2106.04).
determine that the first local device is not trusted in response to not receiving the first local version of the global model from the first local device by the time stamp.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves comparing the occurrence of an event with a specified time condition and determining a trust status based on whether the event occurred by that time). See (MPEP 2106.04).
Regarding claim 14, dependent upon claim 7, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
further configured, responsive to determining that the second local loss value is more reliable than the global loss value, to communicate a trust signal from the host system to the second local device indicative of the host system requesting the second local version of the global model.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyzed failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Regarding claim 15, dependent upon claim 12, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
associate a time stamp with communicating the trust signal to the second local device; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves associating a temporal identifier with an event or communication). See (MPEP 2106.04).
determine that the second local device is not trusted in response to not receiving the second local version of the global model from the second local device by the time stamp.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves determining whether an event occurred within a specified time and assigning a trust status based on that determination). See (MPEP 2106.04).
Regarding claim 16, In step 2A prong 1,
determine whether to communicate an updated version of the global model to the plurality of local devices based on, for each local device of a subset of the plurality of local devices:
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves reviewing local loss values and deciding whether to communicate an updated model based on these values). See (MPEP 2106.04).
If the claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process, but for the recitation of generic computer components, then it falls within the mental process. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 of the 101-analysis, set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
A non-transitory medium storing instructions executable by a processing device to:
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations which do not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
communicate a global model to a plurality of local devices; and
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity). See MPEP (2106.05(g)).
a respective first quantity of samples used by the local device to train the global model, wherein training the global model by the local device yields a respective local version of the global model; and a respective second quantity of samples used by the local device to test the respective local version of the global model.
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitations which do not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
In Step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the
claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitations (I and III), recite mere application of the abstract idea or mere instructions to implement an abstract idea on a computer are deemed insufficient to transform the judicial exception to a patentable invention because the limitations generally apply the use of a generic computer and/or process with the judicial exception, see MPEP 2106.05(f).
Regarding limitation (II), additional elements considered extra/post solution activity, as analyzed above, are activities that are well-understood routine and conventional, specifically: the courts have recognized the computer functions as well‐understood, routine, and conventional functions.
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
As analyzed above, the additional elements, analyzed above, do not integrate the noted judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore, the claim is directed to an abstract idea.
Regarding claim 17, dependent upon claim 16, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
to determine whether to communicate the updated version of the global model to the plurality of local devices based on, for each local device of the subset of the plurality of local devices, a respective local loss value associated with the respective local version of the global model.
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves reviewing local loss values and deciding whether to communicate an updated model based on these values). See (MPEP 2106.04).
Regarding claim 18, dependent upon claim 17, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
to generate the updated version of the global model based on the local versions of the global model from the subset of the plurality of local devices.
Deemed insufficient to transform the judicial exception to a patentable invention because the limitation is directed to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea and are considered to adding the words “apply it” (or an equivalent) with the judicial exception, See MPEP 2106.05(f).
Limitations directed to using the computer as a tool for implementing an abstract idea cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Regarding claim 19, dependent upon claim 18, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
determine whether to communicate a different updated version of the global model to the plurality of local devices based on, for each respective local device of a different subset of the plurality of local devices:
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves evaluating information associated with a subset of local device and deciding whether to communicate a different version of a model based on that information). See (MPEP 2106.04).
a respective third quantity of samples used by the local device to train the different updated version of the global model, wherein training the updated version of the global model by the local device yields a respective local version of the updated version of the global model; and a respective fourth quantity of samples used by the local device to test the respective local version of the updated version of the global model.
Deemed insufficient to transform the judicial exception to a patentable invention because the limitation is directed to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea and are considered to adding the words “apply it” (or an equivalent) with the judicial exception, See MPEP 2106.05(f).
Limitations directed to using the computer as a tool for implementing an abstract idea cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Regarding claim 20, dependent upon claim 19, and fail to resolve the deficiencies identified above by integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
to generate the different updated version of the global model based on the local versions of the updated version of the global model from the different subset of the plurality of local devices.
Deemed insufficient to transform the judicial exception to a patentable invention because the limitation is directed to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea and are considered to adding the words “apply it” (or an equivalent) with the judicial exception, See MPEP 2106.05(f).
Limitations directed to using the computer as a tool for implementing an abstract idea cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
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 – 2 and 6 are rejected under 35 U.S.C. 103 as being unpatentable over KUMAR et al., Pub. No.: US20230065937A1, in view of Luo et al., Pub. No.: US20200413257A1 and ZHENG et al., Pub. No.: US20220114492A1.
Regarding claim 1, KUMAR teaches: A method for federated learning, comprising: communicating a global model and
(KUMAR, “[0033] … The central server computing device 102 may transmit an initial model to the local client computing devices 104. For example, the central server computing device 102 may transmit to the local client computing devices 104 a global model (e.g., newly initialized or partially trained through previous rounds of federated learning) [communicating a global model].”)
a global loss value to the first local device;
(KUMAR, “[0050] 6. Based on the error transmitted from global model [a global loss value to the first local device], this process may repeat iteratively until convergence, or some other stopping factor is reached (see 322, 324: “Recompute predictions and send until converge”)).”)
receiving a local loss value,
(KUMAR, “[0037] These local client computing devices 104 may employ the global model and compute the error in the predictions. In some cases, where this error is large, the local client computing devices 104 may engage in a feedback mechanism to improve the global model, e.g. by supplying additional context such as the amount of error back to the central server computing device 102 [receiving a local loss value,].”)
wherein the local loss value is based on execution of a local version of the global model
(KUMAR, “[0037] … The central server computing device 102 may then take all the outputs of the local models and try to solve an additional optimization, such as optimizing the profit of the company owning the supermarkets. The central server computing device 102 may then send the global model updates to the local client computing devices 104. These local client computing devices 104 may employ the global model and compute the error in the predictions [wherein the local loss value is based on execution of a local version of the global model]. In some cases, where this error is large, the local client computing devices 104 may engage in a feedback mechanism to improve the global model, e.g. by supplying additional context such as the amount of error back to the central server computing device 102.”)
wherein the local version of the global model is generated using the training samples
(KUMAR, “[0015] According to a first aspect, a method performed by a local client computing device is provided. The method includes: training a local model using data from the local client computing device [wherein the local version of the global model is generated using the training samples], resulting in a local model update; sending the local model update to a central server computing device; receiving from the central server computing device a first updated global model;”)
responsive to a result of the analysis of the local loss value being indicative of the local loss value being more preferred than the global loss value
(KUMAR, “[0037] These local client computing devices 104 may employ the global model and compute the error in the predictions. In some cases, where this error is large, [responsive to a result of the analysis of the local loss value being indicative of the local loss value being more preferred than the global loss value] the local client computing devices 104 may engage in a feedback mechanism to improve the global model, e.g. by supplying additional context such as the amount of error back to the central server computing device 102.”)
KUMAR does not teach:
responsive to a valid trust signal from a first local device that is distinct from the apparatus,
communicating a trust signal to the first local device;
responsive to communication of the trust signal to the first local device,
receiving [ ], an indication of a quantity of training samples, and an indication of a quantity of test samples from the first local device
generated by the first local device, on a local test dataset by the first local device,
analyzing the local loss value based on the quantity of training samples and the quantity of test samples; and
Luo teaches:
responsive to a valid trust signal from a first local device that is distinct from the apparatus,
(Luo, “[0030] A gateway server 110 can be associated with gateway device 102, and capable of authenticating a gateway device for a user as well as for local devices (104-0 to -n) [responsive to a valid trust signal from a first local device that is distinct from the apparatus]. A gateway server 110 can have access to network 106 via communication path 116.”)
communicating a trust signal to the first local device;
(Luo, “[0027] A gateway device 102 can include a memory system 126 that stores local device certificates 128, user information 130, and local device state data 132. Local device certificates 128 can be information to enable secure communication between local devices (104-0 to -n) and known, trustful network destinations (e.g., servers) [communicating a trust signal to the first local device]. In some embodiments, certificates 128 can include encryption data for one or more known encryption protocols.”)
responsive to communication of the trust signal to the first local device,
(Luo, “[0030] A gateway server 110 can be associated with gateway device 102, and capable of authenticating a gateway device for a user as well as for local devices (104-0 to -n) [ responsive to communication of the trust signal to the first local device]. A gateway server 110 can have access to network 106 via communication path 116.”)
Luo and KUMAR are related to the same field of endeavor (i.e.: distributed learning). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of Luo with teachings of KUMAR to provide secure, authenticated communication between the private/local machine learning devices and the network model to improve the security and trustworthiness of the federated learning system. (Luo, Abstract).
KUMAR in view of Luo do not teach:
receiving [ ], an indication of a quantity of training samples, and an indication of a quantity of test samples from the first local device
generated by the first local device, on a local test dataset by the first local device,
analyzing the local loss value based on the quantity of training samples and the quantity of test samples;
ZHENG teaches:
receiving [ ] an indication of a quantity of training samples
(ZHENG, “[0091] For example, data fragments of different training members are spliced from top to bottom or from left to right based on numbers of the training members. Finally, training member i transmits the spliced data fragments to the server. After obtaining the spliced data fragments in the form of ciphertext from the k training members, the server performs summation on the spliced data fragments to obtain a total quantity n of training samples [receiving [ ] an indication of a quantity of training samples].”)
… an indication of a quantity of test samples from the first local device
(ZHENG, “[0035] In some embodiments, the performance parameter acquisition module 250 is further configured to obtain a first performance parameter of a corresponding pre-trained model from a data party corresponding to each pre-trained model; test each pre-trained model based on a test set, and obtain a second performance parameter of each pre-trained model, where the test set includes a plurality of test samples [… an indication of a quantity of test samples from the first local device]; and obtain an overfitting parameter of each pre-trained model based on the first performance parameter and the second performance parameter of each pre-trained model. In some embodiments, the test samples in the test set are from one or more data parties, or the test set is from the task party.”)
generated by the first local device, on a local test dataset by the first local device,
(ZHENG, “[0035] … obtain an overfitting parameter of each pre-trained model based on the first performance parameter and the second performance parameter of each pre-trained model. In some embodiments, the test samples in the test set are from one or more data parties [generated by the first local device, on a local test dataset by the first local device], or the test set is from the task party.”)
analyzing the local loss value based on the quantity of training samples
(ZHENG, “[0022] … The training member holds a training sample or training data, and can locally perform model training based on the data held by the training member [analyzing the local loss value based on the quantity of training samples], upload model training data (for example, a local model parameter or gradient information) to the server, obtain an updated model parameter of a target model from the server, and then continue to locally perform model training based on the updated model parameter of the target model.”)
… and the quantity of test samples;
(ZHENG, “[0035] In some embodiments, the performance parameter acquisition module 250 is further configured to obtain a first performance parameter of a corresponding pre-trained model from a data party corresponding to each pre-trained model; test each pre-trained model based on a test set, and obtain a second performance parameter of each pre-trained model, where the test set includes a plurality of test samples; [… and the quantity of test samples] and obtain an overfitting parameter of each pre-trained model based on the first performance parameter and the second performance parameter of each pre-trained model. In some embodiments, the test samples in the test set are from one or more data parties, or the test set is from the task party.”)
ZHENG, KUMAR and Luo are related to the same field of endeavor (i.e.: distributed learning). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of ZHENG with teachings of KUMAR and Luo to evaluate performance parameters of models from multiple candidate data parties and select target data parties to identify more suitable data parties for further model training. (ZHENG, Abstract).
Regarding claim 2, KUMAR in view of Luo and ZHENG teach the method of claim 1.
KUMAR further teaches: further comprising, responsive to the local loss value from the first local device being more preferred than the global loss value, communicating the local loss value to the first local device and to the second local device as an updated global loss value associated with the updated global model.
(KUMAR, “[0047] 3. The central server computing device 102 sends the updated global model [further comprising, responsive to the local loss value from the first local device being more preferred than the global loss value] to the local client computing devices 104 [to the first local device and to a second local device that is distinct from the apparatus] (see 314, 316: “Resend the updated predictions”). Each local client computing device 104 computes a score based on the global model. For example, each local client computing device 104 may determine an error of its local model [communicating the local loss value to the first local device and to the second local device as an updated global loss value associated with the updated global model].”)
Regarding claim 6, KUMAR in view of Luo and ZHENG teach the method of claim 1.
KUMAR further teaches: wherein the method is performed by a cloud server.
(KUMAR, “[0017] According to a second aspect, a method performed by a central server computing device is provided [wherein the method is performed by a cloud server]. The method includes: receiving from a local client computing device a local model update; training a global model using the local model update, resulting in a first updated global model;.”)
Claim(s) 16 – 20 are rejected under 35 U.S.C. 103 as being unpatentable over KUMAR in view of ZHENG and Hard et al., Pub. No.: US20240095582A1.
Regarding claim 16, KUMAR teaches: A non-transitory medium storing instructions executable by a processing device to:
(KUMAR, “[0020] According to a fourth aspect, a central server computing device is provided, including a memory; and a processor [A non-transitory medium storing instructions executable by a processing device to:].”)
communicate a global model to a plurality of local devices; and
(KUMAR, “[0033] … The central server computing device 102 may transmit an initial model to the local client computing devices 104. For example, the central server computing device 102 may transmit to the local client computing devices 104 a global model (e.g., newly initialized or partially trained through previous rounds of federated learning) [communicate a global model to a plurality of local devices].”)
determine whether to communicate an updated version of the global model to the plurality of local devices based on, for each local device of a subset of the plurality of local devices:
(KUMAR, “[0047] 3. The central server computing device 102 sends the updated global model to the local client computing devices 104 [to the plurality of local devices based on, for each local device of a subset of the plurality of local devices] (see 314, 316: “Resend the updated predictions”). Each local client computing device 104 computes a score based on the global model [determine whether to communicate an updated version of the global model]. For example, each local client computing device 104 may determine an error of its local model.
[0048] 4. Each local client computing device 104 sends the computed score (e.g., the error) as context to the central server computing device 102 (see 318, 320: “Send additional factors to optimize predictions”). Sending this context information may include encoding the information using an auto encoder.”)
used by the local device to train the global model, wherein training the global model by the local device yields a respective local version of the global model; and
(KUMAR, “[0046] The local client computing devices 104 may also include additional information (context). In some embodiments, this context may include the price of each item to be procured. The central server computing device 102 may use the local models and context it has received [used by the local device to train the global model, wherein training the global] to obtain an updated global model [model by the local device yields a respective local version of the global model].”)
KUMAR does not teach:
a respective first quantity of samples
a respective second quantity of samples used by the local device to test the respective local version of the global model.
ZHENG teaches:
a respective first quantity of samples
(ZHENG, “[0091] For example, data fragments of different training members are spliced from top to bottom or from left to right based on numbers of the training members. Finally, training member i transmits the spliced data fragments to the server. After obtaining the spliced data fragments in the form of ciphertext from the k training members, the server performs summation on the spliced data fragments to obtain a total quantity n of training samples [a respective first quantity of samples].”)
ZHENG and KUMAR are related to the same field of endeavor (i.e.: distributed learning). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of ZHENG with teachings of KUMAR to evaluate performance parameters of models from multiple candidate data parties and select target data parties to identify more suitable data parties for further model training. (ZHENG, Abstract).
KUMAR in view of ZHENG do not teach:
a respective second quantity of samples used by the local device to test the respective local version of the global model.
Hard teaches:
a respective second quantity of samples used by the local device to test the respective local version of the global model.
(Hard, “[0010] The one or more criteria may include, for example, a threshold quantity of corresponding updates being received from one or more of the computing devices of the population (e.g., such that any other corresponding updates that are received from one or more of the other computing devices of the population may be utilized in generating and/or updating corresponding historical versions of the global ML model), a threshold quantity of time lapsing since the round of decentralized learning was initiated (e.g., 5 minutes, 10 minutes, 15 minutes, 60 minutes, etc.), and/or other criteria.”)
Hard, KUMAR and ZHENG are related to the same field of endeavor (i.e.: distributed learning). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of Hard with teachings of KUMAR and ZHENG to incorporate delayed or late arriving updates from private models after a federated learning rounds has concluded to improve model utilization and convergence by preventing late client updates from being discarded. (Hard, Abstract).
Regarding claim 17, KUMAR in view of ZHENG and Hard teach the method of claim 16.
KUMAR further teaches: further storing instructions executable to determine whether to communicate the updated version of the global model to the plurality of local devices based on, for each local device of the subset of the plurality of local devices, a respective local loss value associated with the respective local version of the global model.
(KUMAR, “[0046] The local client computing devices 104 may also include additional information (context). In some embodiments, this context may include the price of each item to be procured. The central server computing device 102 may use the local models [for each local device of the subset of the plurality of local devices, a respective local loss value] and context it has received to obtain an updated global model [associated with the respective local version of the global model].”)
Regarding claim 18, KUMAR in view of ZHENG and Hard teach the method of claim 17.
KUMAR further teaches: further storing instructions executable to generate the updated version of the global model based on the local versions of the global model from the subset of the plurality of local devices.
(KUMAR, “[0038] … In some embodiments, context is provided to the central server computing device 102 after being encoded by an auto encoder (see auto encoder 208 in FIG. 2 ). The central server computing device 102 may process this information (context) and assign weights for the global model based on the additional information (e.g., output of auto encoder 208) [to generate the updated version of the global model based on the local versions of the global model from the subset of the plurality of local devices]. Embodiments use auto encoders, for example, to compress information before sending, thereby reducing computational power and bandwidth resources.”)
Regarding claim 19, KUMAR in view of ZHENG and Hard teach the method of claim 18.
ZHENG further teaches: a respective third quantity of samples used by the local device to train the different updated version of the global model,
(ZHENG, “[0091] For example, data fragments of different training members are spliced from top to bottom or from left to right based on numbers of the training members. Finally, training member i transmits the spliced data fragments to the server. After obtaining the spliced data fragments in the form of ciphertext from the k training members, the server performs summation on the spliced data fragments to obtain a total quantity n of training samples [a respective third quantity of samples used by the local device to train the different updated version of the global model].”)
Hard further teaches: determine whether to communicate a different updated version of the global model to the plurality of local devices based on, for each respective local device of a different subset of the plurality of local devices:
(Hard, “[0047] Notably, in some versions of those implementations, and as multiple corresponding historical versions of the global ML model are accumulated through multiple rounds of decentralized learning for updating of the global ML model, the corresponding historical versions of the global ML model that are transmitted to each of the computing devices of the population may be uniformly and randomly selected from among the multiple corresponding historical versions of the global ML model [determine whether to communicate a different updated version of the global model to the plurality of local devices based on, for each respective local device of a different subset of the plurality of local devices].”)
wherein training the updated version of the global model by the local device yields a respective local version of the updated version of the global model; and
(Hard, “[0119] … causing the corresponding historical version of the global ML model to be utilized in one or more subsequent rounds of decentralized learning for further updating of the global ML model; and in response to determining that one or more deployment criteria are satisfied, causing a most recently updated version of the global ML model [wherein training the updated version of the global model by the local device yields a respective local version of the updated version of the global model] to be deployed as a final version of the global ML model.”)
a respective fourth quantity of samples used by the local device to test the respective local version of the updated version of the global model.
(Hard, “[0074] At block 454, the system transmits, to a population of computing devices, primary weights for a primary version of a global ML model (e.g., as described with respect to the decentralized learning engine 162, the computing device identification engine 164, and the ML model distribution engine 172 of FIG. 1 ). At block 456, the system causes each of the computing devices of the population to generate a corresponding update for the primary version of the global ML model via utilization of the primary version of the global ML model at each of the computing devices of the population (e.g., as described with respect to the computing device 120 1 [a respective fourth quantity of samples used by the local device to test the respective local version of the updated version of the global model] and the computing device 120 N of FIG. 1).”)
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Hard with teachings of KUMAR and ZHENG for the same reasons disclosed for claim 16.
Regarding claim 20, KUMAR in view of ZHENG and Hard teach the method of claim 19.
Hard further teaches: further storing instructions executable to generate the different updated version of the global model based on the local versions of the updated version of the global model from the different subset of the plurality of local devices.
(Hard, “[0025] … In some versions of those implementations, the computing device 120 1 and the computing device 120 N may replace, in the corresponding storage, any prior weights for the global ML model with the primary weights for the primary version of the global ML model. Each of the computing device 120 1 and the computing device 120 N [based on the local versions of the updated version of the global model from the different subset of the plurality of local devices] may utilize the primary weights for the primary version of the global ML model [to generate the different updated version of the global model] to generate the corresponding update for the global ML model during the given round of decentralized learning.”)
It would have been obvious to one of ordinary skill in the art before the effective filling date of the present application to combine the teachings of Hard with teachings of KUMAR and ZHENG for the same reasons disclosed for claim 16.
Allowable Subject Matter
Claim(s) 3 – 5 and 7 – 15 are objected to as being dependent upon a rejected base claim and would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims and amended to overcome the rejection under 35 U.S.C. 112(b) and 35 U.S.C. 101 set forth in this Office action. The prior art made of record does not teach, make obvious, or suggest the claim limitations as disclosed in applicant's claims.
Claim 3 recites:
wherein the local loss value from the first local device is a first local loss value, and wherein the method further comprises, in association with the analysis of the first local loss value, determining whether the first local loss value is less than a second loss value from the second local device.
Closest prior arts:
KUMAR et al., Pub. No.: US20230065937A1
KUMAR teaches federated learning system in which a local client trains a local model using its own data and sends the resulting model update to a central server. The client then evaluates an updated global model received from the server by computing a score and comparing it with a threshold. If the global model fails to satisfy the client’s local criteria, the client provides context information to the central server. However, KUMAR does not teach comparing the first local loss value generated by a first local device with a second loss value generated by a second local device to determine whether the first local device or the second local device achieved a lower loss.
ZHENG et al., Pub. No.: US20220114492A1.
ZHENG teaches a federated machine learning training system in which a server coordinates multiple data parties. The server provides a pretraining model to candidate data parties, receives their resulting pre-training models, and evaluates each model using a corresponding performance parameter. Based on the evaluation, the server selects one or more target data parties. However, ZHENG does not teach comparing the first local loss value generated by a first local device with a second loss value generated by a second local device to determine whether the first local device or the second local device achieved a lower loss.
Luo et al., Pub. No.: US20200413257A1.
Luo teaches a secure communication system in which an intermediary device connects to a local device and a server using different communication protocols. The intermediary relays authentication protocols. The intermediary relays authentication communications between the local device and the server and once the server authenticates the local device, it enables secure information to be transmitted for storage in the authenticated device’s secure memory. However, Luo does not teach comparing the first local loss value generated by a first local device with a second loss value generated by a second local device to determine whether the first local device or the second local device achieved a lower loss.
The dependent claim(s): 4 – 5, are allowable because of their dependency to claim 3.
Claim 7 recites:
A host system for federated learning, configured to: communicate a global model to a first local device and a second local device, wherein the first local device and the second local device are distinct from and trusted by the host system;
receive from the first local device, a first local loss value based on execution of a first local version of the global model on a first quantity of samples by the first local device;
receive from a second local device, a second local loss value based on execution of a second local version of the global model on a second quantity of samples by the second local device;
determine whether the first local loss value or the second local loss value is more preferred than a global loss value associated with the global model;
responsive to determining that the first local loss value is more preferred than the global loss value, communicate to the first local device and the second local device: a first updated version of the global model based on the first local version of the global model; and a first updated global loss value based on the first local loss value; and
responsive to determining that the second local loss value is more preferred than the global loss value, communicate to the first local device and the second local device: a second updated version of the global model based on the second local version of the global model; and a second updated global loss value based on the second local loss value.
Closest prior arts:
KUMAR et al., Pub. No.: US20230065937A1.
KUMAR teaches federated learning system in which a local client trains a local model using its own data and sends the resulting model update to a central server. The client then evaluates an updated global model received from the server by computing a score and comparing it with a threshold. If the global model fails to satisfy the client’s local criteria, the client provides context information to the central server. However, KUMAR does not teach a federated learning system in which a host distributes a global model to multiple trusted local devices and receives local loss values reflecting the performance of corresponding local versions of the global model. The host compares each local loss value with a global loss value to identify whether a local model provides a more preferred loss. When the first local loss value is more preferred than the global loss value, the host updates the global model based on the first local version of the global model and communicates the resulting first updated global model and a first updated global loss value based on the first local loss value to both the first and second local devices. If the second local loss value is more preferred, the host updates and distributes the global model based on the second local version.
ZHENG et al., Pub. No.: US20220114492A1.
ZHENG teaches a federated machine learning training system in which a server coordinates multiple data parties. The server provides a pretraining model to candidate data parties, receives their resulting pre-training models, and evaluates each model using a corresponding performance parameter. Based on the evaluation, the server selects one or more target data parties. However, ZHENG does not teach a federated learning system in which a host distributes a global model to multiple trusted local devices and receives local loss values reflecting the performance of corresponding local versions of the global model. The host compares each local loss value with a global loss value to identify whether a local model provides a more preferred loss. When the first local loss value is more preferred than the global loss value, the host updates the global model based on the first local version of the global model and communicates the resulting first updated global model and a first updated global loss value based on the first local loss value to both the first and second local devices. If the second local loss value is more preferred, the host updates and distributes the global model based on the second local version.
Luo et al., Pub. No.: US20200413257A1.
Luo teaches a secure communication system in which an intermediary device connects to a local device and a server using different communication protocols. The intermediary relays authentication protocols. The intermediary relays authentication communications between the local device and the server and once the server authenticates the local device, it enables secure information to be transmitted for storage in the authenticated device’s secure memory. However, Luo does not teach a federated learning system in which a host distributes a global model to multiple trusted local devices and receives local loss values reflecting the performance of corresponding local versions of the global model. The host compares each local loss value with a global loss value to identify whether a local model provides a more preferred loss. When the first local loss value is more preferred than the global loss value, the host updates the global model based on the first local version of the global model and communicates the resulting first updated global model and a first updated global loss value based on the first local loss value to both the first and second local devices. If the second local loss value is more preferred, the host updates and distributes the global model based on the second local version.
The dependent claim(s): 8 – 15, are allowable because of their dependency to claim 7.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Nitta et al., Pub. No.: US20240005172A1.
Nitta teaches a plurality of local devices and a server. Each of the local devices includes a processor. The processor selects a mini batch from local data. The processor trains a local model using the mini batch. The processor generates local data relating to the local data included in the mini-batch and indicates information different from a label.
PEZESHKI et al., Pub. No.: US20220116764A1
PEZESHKI teaches transmitting a machine learning model to several user equipment (UEs). The apparatus can receive from each of the UEs, a machine learning processing capability report. The apparatus can also group several UEs in accordance with the machine learning processing capability reports, to receive gradient updates to the machine learning model.
Any inquiry concerning this communication or earlier communications from the examiner
should be directed to MATIYAS T MARU whose telephone number is (571)270-0902 or via email: matiyas.maru@uspto.gov. The examiner can normally be reached Monday 8:00am - Friday 4:00pm EST.
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,
Michelle Bechtold can be reached on (571)431-0762. 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.
/M.T.M./Examiner, Art Unit 2148 /MOHAMED ABOU EL SEOUD/Primary Examiner, Art Unit 2148
Prosecution Insights