DETAILED ACTION
This action is responsive to the Application filed on 11/01/2023. Claims 1-20 are pending in the case. Claims 1, 9 and 16 are independent claims.
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 .
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 07/16/2024 containing 50 NPL is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
The information disclosure statement (IDS) submitted on 07/16/2024 containing 35 NPL is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
The information disclosure statement (IDS) submitted on 10/07/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 19 rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. Claim 19 is dependent from claim 19. Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements. For the purposes of examining the Examiner will interpret the claim to be dependent from claim 16.
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.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Regarding Claim 1:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identify a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating and using judgement to follow conditions down a tree to form a path. See 2106.04.(a)(2).III.C.
The claim recites and determine a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing a set of values and using judgment to map the values. See 2106.04.(a)(2).III.C.
The claim recites for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites and compute…the score for the input data record based on the respective leaves identified which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
at least one processor; non-transitory computer-readable medium; and program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor such that the computing platform is configured to which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f)))
receive a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
input the group of actual parameters into the trained data science model which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
wherein the trained data science model comprises an ensemble of decision trees which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is symmetric which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is configured to receive a respective subset of the features as input which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
via the trained data science model which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) and (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (b) and (c) recites receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
Additional elements (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely links the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)).
The additional element(s) (a) – (g) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 2:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identify at least one a reason code for the score based on the respective overall contribution values for the individual features in the set of features which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using judgement to interpret information and using evaluation to choosing a reason for an outcome/score. See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
transmit the score and the at least one reason code in response to the request which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) recites receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible
Regarding Claim 3:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
prior to receiving the request, train the trained data science model against training data that comprises a plurality of training data records recites the Insignificant Extra-Solution Activity of training a neural network with training data(see MPEP §2106.05(g))
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) recites a well understood and conventional practice of training a neural network with a predetermined loss function quoted from A Review on Conventional Machine Learning vs Deep Learning (Page 1, Col. 2, Paragraph 3, “Conventional machine learning algorithms are based on learning of system by training set to develop a trained model”)
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible
Regarding Claim 4:
The rejection of claim 3 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identifying each realizable path from the root of the individual decision tree to each realizable leaf in the individual decision tree, respectively which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using observation to trace paths and judgement to determine each realizable path down a tree. See 2106.04.(a)(2).III.C.
The claim recites for each identified realizable path, computing a respective first probability by dividing a number of the training data records that were scored during the training based on the identified realizable path by a total number of training data records in the training data which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites for each identified realizable path, identifying a respective score to be assigned to input data records scored by the identified realizable path which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using determining a score to represent each path. See 2106.04.(a)(2).III.C.
The claim recites for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to determine multiple similar criterion at the same level. See 2106.04.(a)(2).III.C.
The claim recites identifying subsets of the respective subset of features that the individual decision tree is configured to receive as input which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to determine data meets a criteria. See 2106.04.(a)(2).III.C.
The claim recites for each identified subset of the respective subset of features, identifying a respective group of realizable paths such that for each level of the individual decision tree in which the same splitting criterion for that level is based on a feature included in the identified subset, the respective path and the realizable paths in the respective group have a same path direction from that level to a next level of the individual decision tree which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to group paths with matching directions/path direction. See 2106.04.(a)(2).III.C.
The claim recites for each identified subset of the respective subset of features, computing a sum of the respective first probabilities for each realizable path in the identified subset which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites and for each identified subset of the respective subset of features, computing a marginal path expectation by multiplying the respective score for the respective path by the sum for the identified subset which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
The claim does not contain elements that would warrant a Step 2A Prong 2 analysis.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include any additional element, when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 5:
The rejection of claim 4 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identifying a selected path to be evaluated for realizability which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass a using observation and evaluation for selecting/choosing a path. See 2106.04.(a)(2).III.C.
The claim recites detecting that a first splitting condition for a first edge in the selected path and a second splitting condition for a second edge in the path contradict each other and excluding the selected path from a list of realizable paths which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass a using observation to choose a path based on using logical reasoning to evaluate splitting conditions. See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
The claim does not contain elements that would warrant a Step 2A Prong 2 analysis.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include any additional element, when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 6:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites based on the identifier of the leaf, determining a set of contribution values to which the identifier maps in a data structure, wherein the determined set of contribution values to which the identifier maps in the data structure is the set of respective individual contribution values which, under the broadest reasonable interpretation, covers performance of the limitation in the mind with physical aid such as pen and paper The limitations encompass evaluating a lookup table and retrieving an associated contribution values based on a identifier . See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
receiving an identifier of a leaf selected from a decision tree in the ensemble which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) recites receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 7:
The rejection of claim 6 is incorporated and further claim recites further additional
elements/limitations:
The claim recites prior to receiving the request, generate a respective set of contribution values for each leaf in the ensemble of decision trees populating the data structure with entries that map the leaves in the ensemble of decision trees to the respective sets of contribution values which, under the broadest reasonable interpretation, covers performance of the limitation in the mind with physical aid such as pen and paper. The limitations encompass evaluating a value associate with each leaf of a tree and tracking the values with a table. See 2106.04.(a)(2).III.C.
The claim recites identifying each realizable path from the root of the individual decision tree to each realizable leaf in the individual decision tree, respectively which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using observation to trace paths and judgement to determine each realizable path down a tree. See 2106.04.(a)(2).III.C.
The claim recites for each identified realizable path, computing a respective first probability by dividing a number of the training data records that were scored during the training based on the identified realizable path by a total number of training data records in the training data which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites for each identified realizable path, identifying a respective score to be assigned to input data records scored by the identified realizable path which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using determining a score to represent each path. See 2106.04.(a)(2).III.C.
The claim recites for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to determine multiple similar criterion at the same level. See 2106.04.(a)(2).III.C.
The claim recites identifying subsets of the respective subset of features that the individual decision tree is configured to receive as input which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to determine data meets a criteria. See 2106.04.(a)(2).III.C.
The claim recites for each identified subset of the respective subset of features, identifying a respective group of realizable paths such that for each level of the individual decision tree in which the same splitting criterion for that level is based on a feature included in the identified subset, the respective path and the realizable paths in the respective group have a same path direction from that level to a next level of the individual decision tree which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using evaluation to group paths with matching directions/path direction. See 2106.04.(a)(2).III.C.
The claim recites for each identified subset of the respective subset of features, computing a sum of the respective first probabilities for each realizable path in the identified subset which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites and for each identified subset of the respective subset of features, computing a marginal path expectation by multiplying the respective score for the respective path by the sum for the identified subset which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
The claim does not contain elements that would warrant a Step 2A Prong 2 analysis.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include any additional element, when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 8:
The rejection of claim 2 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
wherein the at least one reason code comprises a model reason code (MRC) or an adverse action reason code (AARC) which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) recites further details concerning the data that is receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding claim 9:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identify a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating and using judgement to follow conditions down a tree to form a path. See 2106.04.(a)(2).III.C.
The claim recites and determine a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing a set of values and using judgment to map the values. See 2106.04.(a)(2).III.C.
The claim recites for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites and compute…the score for the input data record based on the respective leaves identified which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
a non-transitory computer-readable medium, wherein the non-transitory computer- readable medium is provisioned with program instructions that, when executed by at least one processor, cause a computing platform to which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f)))
receive a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
input the group of actual parameters into the trained data science model which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
wherein the trained data science model comprises an ensemble of decision trees which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is symmetric which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is configured to receive a respective subset of the features as input which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
via the trained data science model which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) and (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (b) and (c) recites receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
Additional elements (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely links the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)).
The additional element(s) (a) – (g) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding claim 10:
The rejection of claim 9 is incorporated and further claim recites further additional elements/limitations:
Claim 10 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 2 found in claim 10.
Regarding claim 11:
The rejection of claim 9 is incorporated and further claim recites further additional elements/limitations:
Claim 11 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 3 found in claim 11.
Regarding claim 12:
The rejection of claim 11 is incorporated and further claim recites further additional elements/limitations:
Claim 12 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 4 found in claim 12.
Regarding claim 13:
The rejection of claim 12 is incorporated and further claim recites further additional elements/limitations:
Claim 13 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 5 found in claim 13.
Regarding claim 14:
The rejection of claim 9 is incorporated and further claim recites further additional elements/limitations:
Claim 14 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 6 found in claim 14.
Regarding claim 15:
The rejection of claim 14 is incorporated and further claim recites further additional elements/limitations:
Claim 15 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 7 found in claim 15.
Regarding claim 16:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites identify a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating and using judgement to follow conditions down a tree to form a path. See 2106.04.(a)(2).III.C.
The claim recites determining a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing a set of values and using judgment to map the values. See 2106.04.(a)(2).III.C.
The claim recites for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites and computing…the score for the input data record based on the respective leaves identified which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
a method carried out by a computing platform which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f)))
receiving a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
inputting the group of actual parameters into the trained data science model which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
wherein the trained data science model comprises an ensemble of decision trees which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is symmetric which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
each individual decision tree in the ensemble is configured to receive a respective subset of the features as input which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features which linking the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)))
via the trained data science model which recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) and (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (b) and (c) recites receiving and sending inputs/outputs which is a well-understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, 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))
Additional elements (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely links the use of a judicial exception to a particular technological environment or field of use(see MPEP 2106.05(h)).
The additional element(s) (a) – (g) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding claim 17:
The rejection of claim 16 is incorporated and further claim recites further additional elements/limitations:
Claim 17 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 2 found in claim 17.
Regarding claim 18:
The rejection of claim 16 is incorporated and further claim recites further additional elements/limitations:
Claim 18 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 3 found in claim 18.
Regarding claim 19:
The rejection of claim 16 is incorporated and further claim recites further additional elements/limitations:
Claim 19 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 4 found in claim 19.
Regarding claim 20:
The rejection of claim 19 is incorporated and further claim recites further additional elements/limitations:
Claim 20 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 5 found in claim 20.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-4, 6-12 and 14-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Prokhorenkova et al(“CatBoost: unbiased boosting with categorical features”, henceforth in view of Prokhorenkova) in view of Lundberg et al(“Consistent Individualized Feature Attribution for Tree Ensembles”, henceforth known as Lundberg).
Regarding claim 1:
Prokhorenkova discloses at least one processor; non-transitory computer-readable medium; and program instructions stored on the non-transitory computer-readable medium that are executable by the at least one processor such that the computing platform is configured to(Prokhorenkova, Page 17, Paragraph 6, “We run all experiments on the same machine with Intel Xeon E3-12xx 2.6GHz, 16 cores, 64GB RAM and run all algorithms with 16 threads”)
Prokhorenkova discloses receive a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input(Prokhorenkova, Page 7, Paragraph 3, “Given all the trees constructed, the leaf values of the final model F are calculated by the standard gradient boosting procedure equally for both modes. Training examples i are matched to leaves leaf0(i)…When the final model F is applied to a new example at testing time” where using model F with new testing examples corresponds to receiving a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input as testing examples are matched to leaves of the tree)
Prokhorenkova discloses input the group of actual parameters into the trained data science model(Prokhorenkova, Page 7, Paragraph 3, “When the final model F is applied to a new example at testing time”), wherein the trained data science model comprises an ensemble of decision trees(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors” where implementing gradient boosting with multiple binary decision trees corresponds to a trained data science model comprises an ensemble of decision trees)
Prokhorenkova discloses each individual decision tree in the ensemble is symmetric(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree. Such trees are balanced…”), each individual decision tree in the ensemble is configured to receive a respective subset of the features as input(Prokhorenkova, Page 2, Paragraph 4, CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t”), and within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree” and Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t” where the oblivious decision trees having the same splitting criterion used is used across an entire level of the tree and the splitting criterion being attributes that identify some feature xk exceeding a threshold corresponds to within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features)
Prokhorenkova discloses for each individual decision tree in the ensemble: identify a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space Rm into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t…Each final region (leaf of the tree) is assigned to a value, which is an estimate of the response y in the region for the regression task or the predicted class label in the case of classification problem” where the decision tree evaluating splitting criterion being based on thresholds down to the leaf and having a value assigned to the leaf corresponds to identifying a leaf that satisfies actual parameters a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree)
Prokhorenkova discloses compute, via the trained data science model, the score for the input data record(Prokhorenkova, Page 16, Algorithm 3, Line 15, where the return statement uses x as the input feature vector) based on the respective leaves identified(Prokhorenkova, Page 15 , Paragraph 5, “We use Function GetLeaf(x,T,σr) to describe how examples are matched to leaves…Given an example with features x…and then choose the leaf of tree T corresponding to features x”)
Prokhorenkova does not discloses determine a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features; for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature.
Lundberg discloses determine a set of respective individual contribution values for the respective leaf(Lundberg, Page 2, Col. 1, Paragraph 2, “Individualized methods that compute feature importance values for a single prediction are less established for trees”), wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features(Lundberg, Page 1, Col. 1, Abstract, “We propose a rich visualization of individualized feature attributions…” (See also Lundberg, Page 3, Col. 1, Paragraph 4, “Feature attribution values are commonly used to identify which features influenced a model’s prediction the most.”))
Lundberg discloses for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature(Lundberg, Page 3, Col. 1, Paragraph 2, “This class covers methods that explain a model’s output as a sum of real values attributed to each input feature.” where the output of a model being the sum of real values attributed to each input feature corresponds to computing a respective overall contribution value based on a sum of individual contributions of each individual feature (See also Lundberg, Page 3, Figure 2, “SHAP (SHapley Additive exPlanation) values explain the output of a function f as a sum of the effects ϕi of each feature being introduced into a conditional expectation”))
References Prokhorenkova and Lundberg are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova and Lundberg before him or her, to use the CatBoost efficient boosted-tree architecture of Prokhorenkova with the Tree SHAP path/subset/probability individualized feature attribution and training-sample cover of Lundberg to enhance the explainability of the models output. The suggestion/motivation for doing so would have been Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model” and Lundberg, Page 1, Col. 1, Paragraph 1, “Understanding why a model made a prediction is important for trust, actionability, accountability, debugging, and many other tasks.”
Regarding claim 2:
The rejection of claim 1 with prior art Prokhorenkova-Lundberg is incorporated and further:
Lundberg further discloses identify at least one a reason code for the score based on the respective overall contribution values for the individual features in the set of features and transmit the score and the at least one reason code in response to the request(Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model…by allocating the difference between the expected value of the model’s output and the current output” where computing the individualized attributions corresponds to identifying at least one a reason code for the score based on the respective overall contribution values for the individual features in the set of features and computing the output/prediction and the matching individualized attributions corresponds to transmitting the score and the at least one reason code in response to the request)
Regarding claim 3:
The rejection of claim 1 with prior art Prokhorenkova-Lundberg is incorporated and further:
Prokhorenkova further discloses prior to receiving the request, train the trained data science model against training data that comprises a plurality of training data records. (Prokhorenkova, Page 16, algorithm 3, where algorithm 3 described building and training the model with the Line 15 return returning the trained prediction function that is used after training for inference/scoring which corresponds to prior to receiving the request, train the trained data science model against training data that comprises a plurality of training data records )
Regarding claim 4:
The rejection of claim 3 with prior art Prokhorenkova-Lundberg is incorporated and further:
Prokhorenkova further discloses for each identified realizable path, identifying a respective score to be assigned to input data records scored by the identified realizable path(Prokhorenkova, Page 2, Paragraph 4, “Each final region (leaf of the tree) is assigned to a value… a decision tree h can be written as…
h
(
x
)
=
∑
j
=
1
J
b
j
1
{
x
∈
R
j
}
,
…where Rj are the disjoint regions corresponding to the leaves of the tree” where each j identifies the current tree node/realizable path being evaluated and the leaf value bj corresponds to a respective score to be assigned to h(x) and a path through splits in the tree and ending in a leaf j of a tree corresponds a realizable path with a leaf value bj and where input record x is evaluated by h(x) to receive bj when
x
∈
R
j
corresponds to identifying a respective score to be assigned to input data records scored by the identified realizable path)
Prokhorenkova further discloses for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree.” and Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t” where the oblivious decision trees having the same splitting criterion used is used across an entire level of the tree and the splitting criterion being attributes that identify some feature xk exceeding a threshold corresponds to for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based )
Lundberg further teaches identifying each realizable path from the root of the individual decision tree to each realizable leaf in the individual decision tree, respectively(Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree” (See also Lundberg, Page 3, Col. 2, Paragraph 4, “The vectors a and b represent the left and right node indexes for each internal node” and Lundberg, Page 4, Col. 1, Algorithm 1 where Algorithm 1 shows the traversal of paths))
Lundberg further discloses for each identified realizable path, computing a respective first probability by dividing a number of the training data records that were scored during the training based on the identified realizable path by a total number of training data records in the training data(Lundberg, Page 4, Col. 1, Algorithm 1, “return G(aj, wraj/rj) + G(bj, wrbj/rj)” where the division of the numbers of training sample reaching the left child raj and right child rbj and the total number of data samples in the current tree node j of subtree rj multiplied by weight w corresponds to computing a respective first probability by dividing a number of training data records scored during training based on the identified realizable path by a total number of training records in the training data as the return yields a weighted expected probability of each child based on training information)
Lundberg further discloses identifying subsets of the respective subset of features that the individual decision tree is configured to receive as input(Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree” (See Also Lundberg, Page 3, Col. 1, Paragraph 7, E[f(x) | xS] is the expected value of the function conditioned on a subset S of the input features))
Lundberg further discloses for each identified subset of the respective subset of features, identifying a respective c paths such that, for each level of the individual decision tree in which the same splitting criterion for that level is based on a feature included in the identified subset, the respective path and the realizable paths in the respective group have a same path direction from that level to a next level of the individual decision tree(Lundberg, Page 2, Figure 1, Model A, where the each trace from one node to another in Model A shows each decision path which are identified realizable paths and each identified path has a subset of the features (Fever and Cough) with an identified group of realizable paths for each level of the decision tree with the same splitting criterion (Fever for the root and Cough for the level below Fever) and wherein each path has the same path direction from one level to the next level)
Lundberg further discloses for each identified subset of the respective subset of features, computing a sum of the respective first probabilities for each realizable path in the identified subset(Lundberg, Page 4, Col. 2, Algorithm 2, Line 7, “w = sum(UNWIND(m,i).w)” where UNWIND is used to calculate the sum of set of weighting terms/path probabilities corresponds to computing the sum of the respective first probabilities for each realizable path in the identified subset (See also Lundberg, Page 4, Col. 1, Paragraph 2, “In Algorithm 2…w which is used to hold the proportion of sets of a given cardinality that are present” and Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree”)) and for each identified subset of the respective subset of features, computing a marginal path expectation by multiplying the respective score for the respective path by the sum for the identified subset(Lundberg, Page 4, Col. 2, Algorithm 2, Lines 8, “ϕmi =ϕmi + w(mi.o − mi.z)vj” where vj being a node when j is a leaf and leaves have a score/output correspond to a respective score, w being sum of set of weighting terms/path probabilities corresponds to a sum for the identified subset and the multiplication of w and vj corresponds to calculate ϕmi corresponds to computing a marginal path expectation as it is the weighted expectation contribution associated with a feature/path configuration).
References Prokhorenkova and Lundberg are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova and Lundberg before him or her, to modify the ML prediction of Prokhorenkova to include the individualized feature attribution of Lundberg to enhance the explainability of the models output. The suggestion/motivation for doing so would have been Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model” and Lundberg, Page 1, Col. 1, Paragraph 1, “Understanding why a model made a prediction is important for trust, actionability, accountability, debugging, and many other tasks.”
Regarding claim 6:
The rejection of claim 1 with prior art Prokhorenkova-Lundberg is incorporated and further:
Lundberg further discloses receiving an identifier of a leaf selected from a decision tree in the ensemble and based on the identifier of the leaf, determining a set of contribution values to which the identifier maps in a data structure(Lundberg, Page 4, Col. 1, Algorithm 1, the recursive helper function that evaluates the contribution of the subtree rooted at node j (procedure G(j, w)) corresponds to receiving an identifier of a leaf selected from a decision tree in the ensemble and based on the identifier of the leaf, determining a set of contribution values to which the identifier maps in a data structure as “return w ·vj” where vj is the value of stored at node j with node j being a leaf (as the if statement checks if it is internal or a leaf) and “return w ·vj” ), wherein the determined set of contribution values to which the identifier maps in the data structure is the set of respective individual contribution values(Lundberg, Page 3, Col. 2, Paragraphs 3-4, “Here we focus on tree models and propose fast SHAP value estimation methods specific to trees and ensembles of trees…we can compute the SHAP values for a tree by estimating E[f (x) | xS] and then using Equation 2 where fx(S) = E[f (x) | xS]” where computing SHAP values for a tree using Algorithm 1 and Equation 2 corresponds to determining a set of contribution values to which the identifier maps in the data structure is the set of respective individual contribution values as SHAP values determined are individuated feature attribution values and Algorithm 1 shows that SHAP values are computed using tree traversal to leaves with leaf values being used in contribution to computation (See also Lundberg, Page 1, Col. 1, Abstract, “To address it we turn to recent applications of game theory and develop fast exact tree solutions for SHAP(SHapley Additive exPlanation) values, which are the unique consistent and locally accurate attribution values” and Lundberg, Page 3, Col. 2, Paragraph 1,“ “SHAP values combine these conditional expectations with the classic Shapley values from game theory to attribute ϕi values to each feature”)
Regarding claim 7:
The rejection of claim 6 with prior art Prokhorenkova-Lundberg is incorporated and further:
Lundberg discloses prior to receiving the request, generate a respective set of contribution values for each leaf(Lundberg, Page 4, Algorithm 1, “if vj ≠ internal then return w ·vj”) in the ensemble of decision trees and populating the data structure(Lundberg, Page 4, Algorithm 2, “ϕ =array of len(x) zeros”) with entries that map the leaves in the ensemble of decision trees to the respective sets of contribution values(Lundberg, Page 4, Algorithm 2, “ϕmi =ϕmi +w(mi.o −mi.z)vj” where algorithm 2 shows the recursive tracking of associated leaf-reaching paths with weighted contribution calculations corresponding to mapping the leafs in the ensemble decision tree to the respective sets of contributing values as each leaf reached during recursion contributes updates to the corresponds ϕ using that leaf’s path m and value vj)
Lundberg discloses determine a set of respective individual contribution values for the respective leaf(Lundberg, Page 2, Col. 1, Paragraph 2, “Individualized methods that compute feature importance values for a single prediction are less established for trees”), wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features(Lundberg, Page 1, Col. 1, Abstract, “We propose a rich visualization of individualized feature attributions…” (See also Lundberg, Page 3, Col. 1, Paragraph 4, “Feature attribution values are commonly used to identify which features influenced a model’s prediction the most.”))
Lundberg discloses for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature(Lundberg, Page 3, Col. 1, Paragraph 2, “This class covers methods that explain a model’s output as a sum of real values attributed to each input feature.” where the output of a model being the sum of real values attributed to each input feature corresponds to computing a respective overall contribution value based on a sum of individual contributions of each individual feature (See also Lundberg, Page 3, Figure 2, “SHAP (SHapley Additive exPlanation) values explain the output of a function f as a sum of the effects ϕi of each feature being introduced into a conditional expectation”))
Lundberg discloses identifying each realizable path from the root of the individual decision tree to each realizable leaf in the individual decision tree, respectively(Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree” (See also Lundberg, Page 3, Col. 2, Paragraph 4, “The vectors a and b represent the left and right node indexes for each internal node” and Lundberg, Page 4, Col. 1, Algorithm 1 where Algorithm 1 shows the traversal of paths))
Lundberg discloses for each identified realizable path, computing a respective first probability by dividing a number of the training data records that were scored during the training based on the identified realizable path by a total number of training data records in the training data(Lundberg, Page 4, Col. 1, Algorithm 1, “return G(aj, wraj/rj) + G(bj, wrbj/rj)” where the division of the numbers of training sample reaching the left child raj and right child rbj and the total number of data samples in the current tree node j of subtree rj multiplied by weight w corresponds to computing a respective first probability by dividing a number of training data records scored during training based on the identified realizable path by a total number of training records in the training data as the return yields a weighted expected probability of each child based on training information)
Lundberg discloses for each identified realizable path, identifying a respective score to be assigned to input data records scored by the identified realizable path(Prokhorenkova, Page 2, Paragraph 4, “Each final region (leaf of the tree) is assigned to a value” where the path ends at a leaf and each leaf have a score/value corresponds to for each path, identify a respective score to be assigned)
Lundberg discloses for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree.” and Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t” where the oblivious decision trees having the same splitting criterion used is used across an entire level of the tree and the splitting criterion being attributes that identify some feature xk exceeding a threshold corresponds to for each level of the individual decision tree, identifying the same feature on which the same splitting criterion specified by the internal nodes at that level is based )
Lundberg discloses identifying subsets of the respective subset of features that the individual decision tree is configured to receive as input(Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree” (See Also Lundberg, Page 3, Col. 1, Paragraph 7, E[f(x) | xS] is the expected value of the function conditioned on a subset S of the input features))
Lundberg discloses for each identified subset of the respective subset of features, identifying a respective group of realizable paths such that, for each level of the individual decision tree in which the same splitting criterion for that level is based on a feature included in the identified subset, the respective path and the realizable paths in the respective group have a same path direction from that level to a next level of the individual decision tree(Lundberg, Page 2, Figure 1, Model A, where the each trace from one node to another in Model A shows each decision path which are identified realizable paths and each identified path has a subset of the features (Fever and Cough) with an identified group of realizable paths for each level of the decision tree with the same splitting criterion (Fever for the root and Cough for the level below Fever) and wherein each path has the same path direction from one level to the next level)
Lundberg discloses for each identified subset of the respective subset of features, computing a sum of the respective first probabilities for each realizable path in the identified subset(Lundberg, Page 4, Col. 2, Algorithm 2, Line 7, “w =sum(UNWIND(m,i).w)” where UNWIND is used to calculate the sum of set of weighting terms/path probabilities corresponds to computing the sum of the respective first probabilities for each realizable path in the identified subset (See also Lundberg, Page 4, Col. 1, Paragraph 2, “In Algorithm 2…w which is used to hold the proportion of sets of a given cardinality that are present” and Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree”))
Lundberg discloses for each identified subset of the respective subset of features, computing a marginal path expectation by multiplying the respective score for the respective path by the sum for the identified subset(Lundberg, Page 4, Col. 2, Algorithm 2, Lines 8, “ϕmi =ϕmi + w(mi.o − mi.z)vj” where vj being a node when j is a leaf and leaves have a score/output correspond to a respective score, w being sum of set of weighting terms/path probabilities corresponds to a sum for the identified subset and the multiplication of w and vj corresponds to calculate ϕmi corresponds to computing a marginal path expectation as it is the weighted expectation contribution associated with a feature/path configuration).
References Prokhorenkova and Lundberg are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova and Lundberg before him or her, to modify the ML prediction of Prokhorenkova to include the individualized feature attribution of Lundberg to enhance the explainability of the models output The suggestion/motivation for doing so would have been Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model” and Lundberg, Page 1, Col. 1, Paragraph 1, “Understanding why a model made a prediction is important for trust, actionability, accountability, debugging, and many other tasks.”
Regarding claim 8:
The rejection of claim 2 with prior art Prokhorenkova-Lundberg is incorporated and further:
Lundberg discloses wherein the at least one reason code comprises a model reason code (MRC) or an adverse action reason code (AARC)(Lundberg, page 4, Col. 2, Algorithm 2, “return ϕ” where the returning of individualized feature attribution values (SHAP values) corresponds to transmitting a reason code comprising a model reason code as the individualized feature attribution explain the prediction of the model (See Also Figure 1, Page 2, “…the individualized attributions explain a single prediction of the model”))
Regarding claim 9:
Prokhorenkova discloses a non-transitory computer-readable medium, wherein the non-transitory computer- readable medium is provisioned with program instructions that, when executed by at least one processor to(Prokhorenkova, Page 17, Paragraph 6, “We run all experiments on the same machine with Intel Xeon E3-12xx 2.6GHz, 16 cores, 64GB RAM and run all algorithms with 16 threads”):
Prokhorenkova discloses receive a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input(Prokhorenkova, Page 7, Paragraph 3, “Given all the trees constructed, the leaf values of the final model F are calculated by the standard gradient boosting procedure equally for both modes. Training examples i are matched to leaves leaf0(i)…When the final model F is applied to a new example at testing time” where using model F with new testing examples corresponds to receiving a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input as testing examples are matched to leaves of the tree)
Prokhorenkova discloses input the group of actual parameters into the trained data science model(Prokhorenkova, Page 7, Paragraph 3, “When the final model F is applied to a new example at testing time”), wherein the trained data science model comprises an ensemble of decision trees(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors” where implementing gradient boosting with multiple binary decision trees corresponds to a trained data science model comprises an ensemble of decision trees)
Prokhorenkova discloses each individual decision tree in the ensemble is symmetric(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree. Such trees are balanced…”), each individual decision tree in the ensemble is configured to receive a respective subset of the features as input(Prokhorenkova, Page 2, Paragraph 4, CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t”), and within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree” and Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t” where the oblivious decision trees having the same splitting criterion used is used across an entire level of the tree and the splitting criterion being attributes that identify some feature xk exceeding a threshold corresponds to within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features)
Prokhorenkova discloses for each individual decision tree in the ensemble: identify a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space Rm into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t…Each final region (leaf of the tree) is assigned to a value, which is an estimate of the response y in the region for the regression task or the predicted class label in the case of classification problem” where the decision tree evaluating splitting criterion being based on thresholds down to the leaf and having a value assigned to the leaf corresponds to identifying a leaf that satisfies actual parameters a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree)
Prokhorenkova discloses compute, via the trained data science model, the score for the input data record(Prokhorenkova, Page 16, Algorithm 3, Line 15, where the return statement uses x as the input feature vector) based on the respective leaves identified(Prokhorenkova, Page 15 , Paragraph 5, “We use Function GetLeaf(x,T,σr) to describe how examples are matched to leaves…Given an example with features x…and then choose the leaf of tree T corresponding to features x”)
Prokhorenkova does not discloses determine a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features; for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature;
Lundberg discloses determine a set of respective individual contribution values for the respective leaf(Lundberg, Page 2, Col. 1, Paragraph 2, “Individualized methods that compute feature importance values for a single prediction are less established for trees”), wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features(Lundberg, Page 1, Col. 1, Abstract, “We propose a rich visualization of individualized feature attributions…” (See also Lundberg, Page 3, Col. 1, Paragraph 4, “Feature attribution values are commonly used to identify which features influenced a model’s prediction the most.”))
Lundberg discloses for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature(Lundberg, Page 3, Col. 1, Paragraph 2, “This class covers methods that explain a model’s output as a sum of real values attributed to each input feature.” where the output of a model being the sum of real values attributed to each input feature corresponds to computing a respective overall contribution value based on a sum of individual contributions of each individual feature (See also Lundberg, Page 3, Figure 2, “SHAP (SHapley Additive exPlanation) values explain the output of a function f as a sum of the effects ϕi of each feature being introduced into a conditional expectation”))
References Prokhorenkova and Lundberg are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova and Lundberg before him or her, to modify the ML prediction of Prokhorenkova to include the individualized feature attribution of Lundberg to enhance the explainability of the models output. The suggestion/motivation for doing so would have been Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model” and Lundberg, Page 1, Col. 1, Paragraph 1, “Understanding why a model made a prediction is important for trust, actionability, accountability, debugging, and many other tasks.”
Regarding claim 10:
The rejection of claim 9 is incorporated in claim 10. Claim 10 is rejected under the same rationale as set forth in the rejection of claim 2.
Regarding claim 11:
The rejection of claim 9 is incorporated in claim 11. Claim 11 is rejected under the same rationale as set forth in the rejection of claim 3.
Regarding claim 12:
The rejection of claim 11 is incorporated in claim 12. Claim 12 is rejected under the same rationale as set forth in the rejection of claim 4.
Regarding claim 14:
The rejection of claim 9 is incorporated in claim 14. Claim 14 is rejected under the same rationale as set forth in the rejection of claim 6.
Regarding claim 15:
The rejection of claim 14 is incorporated in claim 15. Claim 15 is rejected under the same rationale as set forth in the rejection of claim 7.
Regarding claim 16:
Prokhorenkova discloses A method carried out by a computing platform(Prokhorenkova, Page 17, Paragraph 6, “We run all experiments on the same machine with Intel Xeon E3-12xx 2.6GHz, 16 cores, 64GB RAM and run all algorithms with 16 threads”)
Prokhorenkova discloses receiving a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input(Prokhorenkova, Page 7, Paragraph 3, “Given all the trees constructed, the leaf values of the final model F are calculated by the standard gradient boosting procedure equally for both modes. Training examples i are matched to leaves leaf0(i)…When the final model F is applied to a new example at testing time” where using model F with new testing examples corresponds to receiving a request to compute a score for an input data record, the input data record comprising a group of actual parameters that map to a set of features that a trained data science model is configured to receive as input as testing examples are matched to leaves of the tree)
Prokhorenkova discloses inputting the group of actual parameters into the trained data science model(Prokhorenkova, Page 7, Paragraph 3, “When the final model F is applied to a new example at testing time”), wherein the trained data science model comprises an ensemble of decision trees(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors” where implementing gradient boosting with multiple binary decision trees corresponds to a trained data science model comprises an ensemble of decision trees)
Prokhorenkova discloses each individual decision tree in the ensemble is symmetric(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree. Such trees are balanced…”), each individual decision tree in the ensemble is configured to receive a respective subset of the features as input(Prokhorenkova, Page 2, Paragraph 4, CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t”), and within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features(Prokhorenkova, Page 6, Paragraph 3, “In CatBoost, base predictors are oblivious decision trees…also called decision tables…oblivious means that the same splitting criterion is used across an entire level of the tree” and Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space…into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t” where the oblivious decision trees having the same splitting criterion used is used across an entire level of the tree and the splitting criterion being attributes that identify some feature xk exceeding a threshold corresponds to within each individual decision tree, internal nodes that are positioned in a same level designate a same splitting criterion based on a same feature selected from the respective subset of features)
Prokhorenkova discloses for each individual decision tree in the ensemble: identifying a respective leaf such that the actual parameters satisfy a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree to the respective leaf(Prokhorenkova, Page 2, Paragraph 4, “CatBoost is an implementation of gradient boosting, which uses binary decision trees as base predictors. A decision tree…is a model built by a recursive partition of the feature space Rm into several disjoint regions (tree nodes) according to the values of some splitting attributes α. Attributes are usually binary variables that identify that some feature xk exceeds some threshold t…Each final region (leaf of the tree) is assigned to a value, which is an estimate of the response y in the region for the regression task or the predicted class label in the case of classification problem” where the decision tree evaluating splitting criterion being based on thresholds down to the leaf and having a value assigned to the leaf corresponds to identifying a leaf that satisfies actual parameters a series of splitting conditions for edges that connect nodes in a respective path from a root of the individual decision tree)
Prokhorenkova discloses computing, via the trained data science model, the score for the input data record (Prokhorenkova, Page 16, Algorithm 3, Line 15, where the return statement uses x as the input feature vector) based on the respective leaves identified(Prokhorenkova, Page 15 , Paragraph 5, “We use Function GetLeaf(x,T,σr) to describe how examples are matched to leaves…Given an example with features x…and then choose the leaf of tree T corresponding to features x”)
Prokhorenkova does not discloses determining a set of respective individual contribution values for the respective leaf, wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features; for each individual feature in the set of features, computing a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature
Lundberg discloses determine a set of respective individual contribution values for the respective leaf(Lundberg, Page 2, Col. 1, Paragraph 2, “Individualized methods that compute feature importance values for a single prediction are less established for trees”), wherein each of the respective individual contribution values maps to a respective feature found in the respective subset of features(Lundberg, Page 1, Col. 1, Abstract, “We propose a rich visualization of individualized feature attributions…” (See also Lundberg, Page 3, Col. 1, Paragraph 4, “Feature attribution values are commonly used to identify which features influenced a model’s prediction the most.”))
Lundberg discloses for each individual feature in the set of features, compute a respective overall contribution value based on a sum of the respective individual contribution values that map to that individual feature(Lundberg, Page 3, Col. 1, Paragraph 2, “This class covers methods that explain a model’s output as a sum of real values attributed to each input feature.” where the output of a model being the sum of real values attributed to each input feature corresponds to computing a respective overall contribution value based on a sum of individual contributions of each individual feature (See also Lundberg, Page 3, Figure 2, “SHAP (SHapley Additive exPlanation) values explain the output of a function f as a sum of the effects ϕi of each feature being introduced into a conditional expectation”))
References Prokhorenkova and Lundberg are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova and Lundberg before him or her, to modify the ML prediction of Prokhorenkova to include the individualized feature attribution of Lundberg to enhance the explainability of the models output. The suggestion/motivation for doing so would have been Lundberg, Page 2, Figure 1, “The individualized attributions explain a single prediction of the model” and Lundberg, Page 1, Col. 1, Paragraph 1, “Understanding why a model made a prediction is important for trust, actionability, accountability, debugging, and many other tasks.”
Regarding claim 17:
The rejection of claim 16 is incorporated in claim 17. Claim 17 is rejected under the same rationale as set forth in the rejection of claim 2.
Regarding claim 18:
The rejection of claim 16 is incorporated in claim 18. Claim 18 is rejected under the same rationale as set forth in the rejection of claim 3.
Regarding claim 19:
The rejection of claim 16 is incorporated in claim 19. Claim 19 is rejected under the same rationale as set forth in the rejection of claim 4.
Claim(s) 5, 13 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Prokhorenkova et al(“CatBoost: unbiased boosting with categorical features”, henceforth in view of Prokhorenkova) in view of Lundberg et al(“Consistent Individualized Feature Attribution for Tree Ensembles”, henceforth known as Lundberg) and further in view of Aytekin(“Neural Networks are Decision Trees”)
Regarding claim 5:
The rejection of claim 4 with prior art Prokhorenkova-Lundberg is incorporated and further:
Lundberg further discloses identifying a selected path to be evaluated for realizability(Lundberg, Page 4, Col. 1, Paragraph 3, “The intuition of the polynomial time algorithm is to recursively keep track of what proportion of all possible subsets flow down into each of the leaves of the tree”)
The Prokhorenkova-Lundberg does not disclose, however Aytekin discloses detecting that a first splitting condition for a first edge in the selected path and a second splitting condition for a second edge in the path contradict each other and excluding the selected path from a list of realizable paths(Aytekin, Page 5, Figure 2, Figure 3 and Col. 2 Paragraph 1, “However, we notice that a lot of the rules in the decision tree is redundant, and hence some paths in the decision tree becomes invalid. An example to redundant rule is checking x < 0.32 after x < −1.16 rule holds” where the model in Figure 2 having paths removed to form the model in Figure 3(a) corresponds to detecting that a first splitting condition for a first edge in the selected path and a second splitting condition for a second edge in the path contradict each other and excluding the selected path from a list of realizable paths))
References Prokhorenkova-Lundberg and Aytekin are analogous art because they are from the same field of endeavor of improving the practical usability of tree ensemble ML systems interpretability of neural models.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Prokhorenkova-Lundberg and Aytekin before him or her, to modify the decision trees of Prokhorenkova-Lundberg to include the removal of redundant paths of Aytekin to provide lossless pruning of obsolete data to improve the model. The suggestion/motivation for doing so would have been Aytekin, Page 3, Col. 1, Paragraph 1, “may occur violating and redundant rules that would provide lossless pruning of the NN-equivalent tree. Another observation is that, it is highly likely that not all categories will be realized during training due to the possibly much larger number of categories(tree leaves) than training data. These categories can be pruned as well based on the application, and the data falling into these categories can be considered invalid”
Regarding claim 13:
The rejection of claim 12 is incorporated in claim 13. Claim 13 is rejected under the same rationale as set forth in the rejection of claim 5.
Regarding claim 20:
The rejection of claim 19 is incorporated in claim 20. Claim 20 is rejected under the same rationale as set forth in the rejection of claim 5.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHARLES JEFFREY JONES JR whose telephone number is (703)756-1414. The examiner can normally be reached Monday - Friday 8:00 - 5:00 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, Kakali Chaki can be reached at 571-272-3719. 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.
/C.J.J./Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights