# US 2021/0174264 A1 · Training tree-based machine-learning modeling algorithms for predicting outputs and generating explanatory data

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0174264 A1
- Document ID: US20210174264A1
- Status: Published application
- Kind: A1
- Application number: 17/182,053
- Filed: 2021-02-22
- Published: 2021-06-10
- Earliest claimed: 2019-04-10
- Assignees: Equifax Inc.
- Applicants: Equifax Inc.
- Inventors: Lewis JORDAN; Matthew TURNER; Finto ANTONY
- Web page: https://patentgrep.com/patent/US20210174264A1

## CPC

- G06N 20/20: Ensemble learning
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 5/04: Inference or reasoning models
- G06N 5/045: Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Certain aspects involve training tree-based machine-learning models for computing predicted responses and generating explanatory data for the models. For example, independent variables having relationships with a response variable are identified. Each independent variable corresponds to an action or observation for an entity. The response variable has outcome values associated with the entity. Splitting rules are used to generate the tree-based model, which includes decision trees for determining relationships between independent variables and a predicted response associated with the response variable. The tree-based model is iteratively adjusted to enforce monotonicity with respect to representative response values of the terminal nodes. For instance, one or more decision trees are adjusted such that one or more representative response values are modified and a monotonic relationship exists between each independent variable and the response variable. The adjusted model is used to output explanatory data indicating relationships between independent variable changes and response variable changes.

## Claims

### Claim 1

(canceled)

### Claim 2

A method comprising:
generating a tree-based machine-learning model that is a memory structure comprising interconnected parent nodes and terminal nodes, wherein each parent node includes a respective splitting variable that causes the parent node to be connected via links to a respective pair of child nodes, wherein the terminal nodes includes respective representative response values based on values of the splitting variables, wherein generating the tree-based machine-learning model comprises:
determining a splitting rule for partitioning data samples in a decision tree;
partitioning, based on the splitting rule, data samples into a first tree region and a second tree region;
computing a first representative response value from the data samples in first tree region and a second representative response value from the data samples in second tree region, wherein a monotonicity constraint for the decision tree is violated by a set of representative response values including the first and second representative response values, a representative response value for a closest lower neighboring region of the first tree region, and a representative response value for a closest upper neighboring region of the second tree region;
applying a modification to the decision tree;
computing, with the decision tree having the modification, a first modified representative response value from data samples partitioned into the first tree region and a second modified representative response value from data samples partitioned into the second tree region, wherein:
the monotonicity constraint is satisfied by a modified set of representative response values, and
the modified set of representative response values includes the first and second modified representative response values, the representative response value for the closest lower neighboring region of the first tree region, and the representative response value for the closest upper neighboring region of the second tree region; and
outputting the decision tree having the modification.

### Claim 3 (depends on 2)

The method of claim 2, further comprising:
computing, based on the decision tree as outputted, explanatory data indicating relationships between (i) changes in a response variable computed with the tree-based machine-learning model and (ii) changes in one or more independent variables represented by the tree-based machine-learning model.

### Claim 4 (depends on 3)

The method of claim 3, wherein the relationships comprise respective contributions of the one or more independent variables to an output value of the response variable.

### Claim 5 (depends on 4)

The method of claim 4, wherein the output value of the response variable comprises a risk assessment.

### Claim 6 (depends on 3)

The method of claim 3, wherein the independent variables comprise a first independent variable and a second independent variable, wherein generating the explanatory data for the first independent variable having an input value comprises:
identifying a first optimizing value of the first independent variable and a second optimizing value of the second independent variable that cause the tree-based machine-learning model to output an optimal value for the response variable;
computing an additional value for the response variable by applying the tree-based machine-learning model to the first independent variable having the input value and the second independent variable having the second optimizing value;
computing a difference between the optimal value for the response variable and the additional value for the response variable; and
outputting the explanatory data for the first independent variable based on the difference.

### Claim 7 (depends on 6)

The method of claim 6, wherein generating the explanatory data for the second independent variable having an additional input value comprises:
computing a third value for the response variable by applying the tree-based machine-learning model to the first independent variable having the first optimizing value and the second independent variable having the additional input value;
computing an additional difference between the optimal value for the response variable and the third value for the response variable; and
excluding, based on the difference being larger than the additional difference, an explanation of the second independent variable from the explanatory data.

### Claim 8 (depends on 7)

The method of claim 7, wherein output values of the response variable comprises risk assessments, respectively.

### Claim 9 (depends on 3)

The method of claim 3, wherein the independent variables comprise a first independent variable having a first input value and a second independent variable having a second input value, wherein generating the explanatory data for the first independent variable having the first input value comprises:
identifying a first optimizing value of the first independent variable and a second optimizing value of the second independent variable that cause the tree-based machine-learning model to output an optimal value for the response variable;
computing a first output value for the response variable by applying the tree-based machine-learning model to the first independent variable having the first optimizing value and the second independent variable having the second input value;
identifying a second output value for the response variable that is computed by applying the tree-based machine-learning model to the first independent variable having the first input value and the second independent variable having the second input value
computing a difference between the first output value and the second output value; and
outputting the explanatory data for the first independent variable based on the difference.

### Claim 10 (depends on 9)

The method of claim 9, wherein generating the explanatory data for the second independent variable having the second input value comprises:
computing a third output value for the response variable by applying the tree-based machine-learning model to the first independent variable having the first input value and the second independent variable having the second optimizing value;
computing an additional difference between the first output value and the third output value; and
excluding, based on the difference being larger than the additional difference, an explanation of the second independent variable from the explanatory data.

### Claim 11 (depends on 10)

The method of claim 10, wherein output values of the response variable comprises risk assessments, respectively.

### Claim 12

A method comprising:
receiving a target dataset having values for predictors, respectively;
generating a risk assessment from the target data by, at least, applying a tree-based machine-learning model to the target dataset; and
generating explanatory data identifying respective contributions of the predictors to the risk assessment, wherein the tree-based machine-learning model complies with a monotonicity constraint that facilitates the generation of the explanatory data.

### Claim 13 (depends on 12)

The method of claim 12, wherein the tree-based machine-learning model complying with the monotonicity constraint comprises:
accessing an input dataset comprising a first input, a second input, a third input, and a fourth input,
applying, to the input dataset, a decision tree of the tree-based machine-learning model, the decision tree having a first tree region, a second tree region, a third tree region, and a fourth tree region, wherein
(a) the first tree region is a closest lower neighboring region of the second tree region,
(b) the second tree region is a left split of a splitting rule,
(c) the third tree region is a right split of the splitting rule, and
(d) the fourth tree region is a closest upper neighboring region of the third tree region; and
computing an output dataset comprising:
(a) a first response value determined by traversing, based on the first input, the decision tree to the first tree region,
(b) a second response value determined by traversing, based on the second input, the decision tree to the second tree region,
(c) a third response value determined by traversing, based on the third input, the decision tree to the third tree region, and
(d) a fourth response value determined by traversing, based on the fourth input, the decision tree to the fourth tree region,
wherein the output dataset is monotonic with respect to the input dataset.

### Claim 14 (depends on 13)

The method of claim 13, wherein the output dataset being monotonic with respect to the input dataset comprises:
(a) the first input is larger than the second input, the second input is larger than the third input, the third input is larger than the fourth input, the first response value is larger than the second response value, the second response value is larger than the third response value, and the third response value is larger than the fourth response value; or
(b) the first input is smaller than the second input, the second input is smaller than the third input, the third input is smaller than the fourth input, the first response value is smaller than the second response value, the second response value is smaller than the third response value, and the third response value is smaller than the fourth response value.

### Claim 15 (depends on 13)

The method of claim 13, wherein the predictors comprise a first predictor and a second predictor, wherein generating the explanatory data for the first predictor having an input value comprises:
identifying a first optimizing value of the first predictor and a second optimizing value of the second predictor that cause the tree-based machine-learning model to output an optimal value for the risk assessment;
computing an additional value for the risk assessment by applying the tree-based machine-learning model to the first predictor having the input value and the second predictor having the second optimizing value;
computing a difference between the optimal value for the risk assessment and the additional value for the response variable; and
outputting the explanatory data for the first predictor based on the difference.

### Claim 16 (depends on 15)

The method of claim 15, wherein generating the explanatory data for the second predictor having an additional input value comprises:
computing a third value for the risk assessment by applying the tree-based machine-learning model to the first predictor having the first optimizing value and the second predictor having the additional input value;
computing an additional difference between the optimal value for the risk assessment and the third value for the risk assessment; and
excluding, based on the difference being larger than the additional difference, an explanation of the second predictor from the explanatory data.

### Claim 17 (depends on 13)

The method of claim 13, wherein the predictors comprise a first predictor having a first input value and a second predictor having a second input value, wherein generating the explanatory data for the first predictor having the first input value comprises:
identifying a first optimizing value of the first predictor and a second optimizing value of the second predictor that cause the tree-based machine-learning model to output an optimal value for the risk assessment;
computing a first output value for the risk assessment by applying the tree-based machine-learning model to the first predictor having the first optimizing value and the second predictor having the second input value;
identifying a second output value for the risk assessment that is computed by applying the tree-based machine-learning model to the first predictor having the first input value and the second predictor having the second input value
computing a difference between the first output value and the second output value; and
outputting the explanatory data for the first predictor based on the difference.

### Claim 18 (depends on 17)

The method of claim 17, wherein generating the explanatory data for the second predictor having the second input value comprises:
computing a third output value for the risk assessment by applying the tree-based machine-learning model to the first predictor having the first input value and the second predictor having the second optimizing value;
computing an additional difference between the first output value and the second output value; and
excluding, based on the difference being larger than the additional difference, an explanation of the second predictor from the explanatory data.

### Claim 19

A non-transitory computer-readable medium having program code stored thereon, wherein the program code is executable by processing hardware to cause the processing hardware to perform operations comprising:
generating a tree-based machine-learning model that is a memory structure comprising interconnected parent nodes and terminal nodes, wherein each parent node includes a respective splitting variable that causes the parent node to be connected via links to a respective pair of child nodes, wherein the terminal nodes includes respective representative response values based on values of the splitting variables, wherein generating the tree-based machine-learning model comprises:
determining a splitting rule for partitioning data samples in a decision tree;
partitioning, based on the splitting rule, data samples into a first tree region and a second tree region;
computing a first representative response value from the data samples in first tree region and a second representative response value from the data samples in second tree region, wherein a monotonicity constraint for the decision tree is violated by a set of representative response values including the first and second representative response values, a representative response value for a closest lower neighboring region of the first tree region, and a representative response value for a closest upper neighboring region of the second tree region;
applying a modification to the decision tree;
computing, with the decision tree having the modification, a first modified representative response value from data samples partitioned into the first tree region and a second modified representative response value from data samples partitioned into the second tree region, wherein:
the monotonicity constraint is satisfied by a modified set of representative response values, and
the modified set of representative response values includes the first and second modified representative response values, the representative response value for the closest lower neighboring region of the first tree region, and the representative response value for the closest upper neighboring region of the second tree region; and
outputting the decision tree having the modification.

### Claim 20 (depends on 19)

The non-transitory computer-readable medium of claim 19, the operations further comprising computing, based on the decision tree as outputted, explanatory data indicating relationships between (i) changes in a response variable computed with the tree-based machine-learning model and (ii) changes in one or more independent variables represented by the tree-based machine-learning model,
wherein the independent variables comprise a first independent variable and a second independent variable,
wherein generating the explanatory data for the first independent variable having an input value comprises:
identifying a first optimizing value of the first independent variable and a second optimizing value of the second independent variable that cause the tree-based machine-learning model to output an optimal value for the response variable;
computing an additional value for the response variable by applying the tree-based machine-learning model to the first independent variable having the input value and the second independent variable having the second optimizing value;
computing a difference between the optimal value for the response variable and the additional value for the response variable; and
outputting the explanatory data for the first independent variable based on the difference.

### Claim 21 (depends on 19)

The non-transitory computer-readable medium of claim 19, the operations further comprising computing, based on the decision tree as outputted, explanatory data indicating relationships between (i) changes in a response variable computed with the tree-based machine-learning model and (ii) changes in one or more independent variables represented by the tree-based machine-learning model,
wherein the independent variables comprise a first independent variable having a first input value and a second independent variable having a second input value,
wherein generating the explanatory data for the first independent variable having the first input value comprises:
identifying a first optimizing value of the first independent variable and a second optimizing value of the second independent variable that cause the tree-based machine-learning model to output an optimal value for the response variable;
computing a first output value for the response variable by applying the tree-based machine-learning model to the first independent variable having the first optimizing value and the second independent variable having the second input value;
identifying a second output value for the response variable that is computed by applying the tree-based machine-learning model to the first independent variable having the first input value and the second independent variable having the second input value
computing a difference between the first output value and the second output value; and
outputting the explanatory data for the first independent variable based on the difference.
