# US 2022/0036385 A1 · Segment Valuation in a Digital Medium Environment

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

- Number: US 2022/0036385 A1
- Document ID: US20220036385A1
- Status: Published application
- Kind: A1
- Application number: 17/503,702
- Filed: 2021-10-18
- Published: 2022-02-03
- Earliest claimed: 2016-11-17
- Assignees: Adobe Inc.
- Applicants: Adobe Inc.
- Inventors: Kourosh Modarresi; Jamie Mark Diner; Elizabeth T. Chin; Aran Nayebi
- Web page: https://patentgrep.com/patent/US20220036385A1

## CPC

- G06Q 30/0201: Market modelling; Market analysis; Collecting market data
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 20/00: Machine learning
- G06N 20/20: Ensemble learning
- G06Q 30/02: Marketing; Price estimation or determination; Fundraising
- G06Q 30/0202: Market predictions or forecasting for commercial activities
- G06Q 30/0269: {based on user profile or attribute}
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Segment valuation techniques usable in a digital medium environment are described. To do so, a segment valuation system first identifies the attributes that are significant in achievement of a desired metric (e.g., conversion) and then values segments based on those significant attributes. Attributes are selected from the trained model based on significance of those attributes towards achieving the desired metric. A valuation of a segment may then be calculated based on the valuations of these attributes. For example, inclusion of the selected attributes within a segment, and the valuations of those selected attributes, is then used by the segment valuation system to generate data describing a value of the segment towards achieving the metric.

## Claims

### Claim 1

In a digital medium segmentation valuation environment, a method implemented by a computing device, the method comprising:
obtaining, by the computing device, data describing a population and achievement of a metric by the population;
processing, by the computing device using a machine learning model, the data by employing a penalty term configured to reduce overfitting and underfitting of the model;
generating, by the computing device using the machine learning model, a valuation of a segment of the population based on a significance of respective attributes of a plurality of attributes of the population on the achievement of the metric; and
outputting, by the computing device, the valuation of the segment of the population.

### Claim 2 (depends on 1)

The method as described in claim 1, further comprising selecting, by the computing device, a subset of attributes from the plurality of attributes, the selection based on the significance, and wherein the generating the valuation is based on the subset.

### Claim 3 (depends on 2)

The method as described in claim 2, wherein the selecting is based on a score generated for each attribute of the plurality of attributes by the machine learning model regarding the significance in achieving the metric.

### Claim 4 (depends on 3)

The method as described in claim 3, wherein the score defines a measure of accuracy in the significance in achieving the metric.

### Claim 5 (depends on 4)

The method as described in claim 4, wherein the score is an F-score that describes significance of respective said attributes towards the achievement of the metric.

### Claim 6 (depends on 1)

The method as described in claim 1, wherein the penalty term is configured to reduce overfitting and underfitting of the model as part of balancing accuracy and complexity of the model, respectively.

### Claim 7 (depends on 1)

The method as described in claim 1, wherein the model is an ensemble model formed using a plurality of sub-models having weighted contributions towards an overall result of the ensemble model to describe the achievement of the metric by the population.

### Claim 8 (depends on 1)

The method as described in claim 1, wherein the machine learning includes a linear model penalized with L 1 normal regularization (Lasso), a random forest, a guided random forest, an adaptive boosting ensemble model (AdaBoost), or gradient boosted trees (GBRT).

### Claim 9

One or more computer readable storage media having instructions stored thereon that, responsive to execution by a processing system, causes the processing system to perform operations comprising:
obtaining a training dataset including a population and achievement of a metric by the population; and
training a machine learning model using the training dataset to generate a trained machine learning model that values a segment of the population based on a significance of respective attributes of a plurality of attributes on the achievement of the metric, the training using a penalty term as part of regularization to reduce overfitting and underfitting of the machine learning model.

### Claim 10 (depends on 9)

One or more computer readable storage media as described in claim 9, wherein the penalty term is configured to reduce overfitting and underfitting of the model as part of balancing accuracy and complexity of the model, respectively.

### Claim 11 (depends on 9)

One or more computer readable storage media as described in claim 9, the operations further comprising selecting a subset of attributes from the plurality of attributes, the selection based on the significance, and wherein the generating the values of the segment is based on the subset.

### Claim 12 (depends on 11)

One or more computer readable storage media as described in claim 11, wherein the selecting is based on a score generated for each attribute of the plurality of attributes by the machine learning model regarding the significance in achieving the metric.

### Claim 13 (depends on 12)

One or more computer readable storage media as described in claim 12, wherein the score defines a measure of accuracy in the significance in achieving the metric.

### Claim 14 (depends on 13)

One or more computer readable storage media as described in claim 13, wherein the score is an F-score that describes significance of respective said attributes towards the achievement of the metric.

### Claim 15

In a digital medium segmentation valuation environment, a system comprising:
means for processing, using a machine learning model, data describing a population and achievement of a metric by the population, the data by employing a penalty term configured to reduce overfitting and underfitting of the model;
means for generating, using the machine learning model, a valuation of a segment of the population based on a significance of respective attributes of a plurality of attributes of the population on the achievement of the metric; and
means for outputting the valuation of the segment of the population.

### Claim 16 (depends on 15)

The system as described in claim 15, further comprising means for selecting a subset of attributes from the plurality of attributes based on the significance of the respective attributes on the achievement of the metric, and wherein the generating means is configured to generate the valuation is based on the subset.

### Claim 17 (depends on 15)

The system as described in claim 15, wherein the selecting means is configured to select the subset based on a score generated for each attribute of the plurality of attributes by the machine learning model regarding the significance in achieving the metric.

### Claim 18 (depends on 17)

The system as described in claim 17, wherein the score defines a measure of accuracy in the significance in achieving the metric.

### Claim 19 (depends on 18)

The system as described in claim 18, wherein the score is an F-score that describes significance of respective said attributes towards the achievement of the metric.

### Claim 20 (depends on 15)

The system as described in claim 15, wherein the penalty term is configured to reduce overfitting and underfitting of the model as part of balancing accuracy and complexity of the model, respectively.