Bulk Import Assets > Template file properties > AI Model
  

AI Model

An AI model asset represents the essential parameters of an AI model. These parameters also include the environment where the model is run, the input parameters and the output data, and the business purpose and rules for the model. AI models provide insights for many activities so that you can make decisions faster and easier or provide prediction modeling.
The following table describes the columns where you must enter asset values for AI models:
Column
Type
Description
Reference ID
Optional
Reference ID is the unique reference identifier for the asset. If you don't specify a reference ID, the system automatically assigns a unique value to the object.
The identifier that you enter here appears in the Reference ID field on the user interface.
The system uses the following format to generate a reference ID:
<Prefix>-<Number>
Example: AI-1, DS-10
If you specify a reference ID, ensure that the value that you enter doesn’t use the format that is used by the system to automatically generate reference IDs.
Type of Entry: Alphanumeric characters, including spaces and special characters. There are no restrictions on special characters.
Example: AST ID-18, SYS REF:98
Name
Required
Name for the asset. To distinguish this asset from other assets, consider entering a unique name.
Type of Entry: Alphanumeric characters, including spaces and special characters. There are no restrictions on special characters.
Example: Model 1, Model Name: 29, System Name
Description
Optional
Definition of the asset. The description appears in the asset page.
Enter a detailed description of the asset to help users understand how the asset is used in a governance context.
Type of Entry: Formatted text.
To add an image to the description, use the following syntax:
<p><img src=""data:image/png;base64,""> <img src=""data:image/png;base64,""></p>
You can add only an image of Base64 format to the description. Other image formats are not supported.
AI Model Purpose
Optional
Describe the business or technical purpose of the asset.
Type of Entry: Formatted text
Architecture Type
Optional
Defines the architecture and algorithms of the AI Model asset.
You can select one of the following options:
  • - Linear Regression. The architecture type that computes the linear relationship between the dependent variable and other independent variables.
  • - Logistic Regression. The architecture model that accomplishes binary classification tasks by predicting the probability of an outcome, event, or observation.
  • - Lineage Discrimant Analysis. The architecture model that is used as a classifier and a dimensionality reduction algorithm.
  • - Decision Trees. The architecture type that models decisions, outcomes, and predictions using a flowchart-like tree structure, commonly used for classification and regression tasks.
  • - K-Nearest Neighbors. The architecture model that uses proximity to form classifications or predictions about the grouping of an individual data point. This architecture is used for solving both classification and regression problems.
  • - Naive Bayes.The architecture model that works on the principle of conditional probability used for predictive modeling and classification tasks.
  • - Learning Vector Quantization. Learning Vector Quantization (LVQ) is an algorithm commonly used for statistical classification and pattern recognition.
  • - Support Vector Machines. The architecture type used for linear or nonlinear classification, regression, and outlier detection tasks, where the algorithm creates a line or a hyperplane which separates the data into classes.
  • - Others. If none of the other options meet your requirements, select this option.
Bias Score
Optional
Measure of bias in the model. A higher value indicates a greater difference between the predicted result and the actual result. Use the bias score to determine if the assumptions in the model are correct.
Type of Entry: Numeric value. Enter a value between 0 and 100. Ensure that the value you specify has a maximum precision of 5 digits and a scale of 2 decimal places.
Drift Score
Optional
Measure of degradation of the model. A higher value indicates greater degradation. Use the drift score to determine if the model can provide results within an acceptable range of accuracy.
Type of Entry: Numeric value. Enter a value between 0 and 100. Ensure that the value you specify has a maximum precision of 5 digits and a scale of 2 decimal places.
Environment
Optional
Select the environment for the asset.
You can select one of the following options:
  • - Training
  • - Validation
  • - Production
Input
Optional
Input parameters for the AI model.
Type of Entry: Formatted text
Libraries
Optional
Software libraries that are used to build the AI model.
Type of Entry: Alphanumeric characters, including spaces and special characters.
Lifecycle
Optional
Lifecycle status of the asset. The default status is Published.
You can select one of the following options:
  • - Draft. Typically used when you want to create a fresh asset.
  • - In Review. Typically used to indicate an asset that is currently under review by stakeholders.
    • Note: Informatica recommends that you don't use a bulk import template to set an asset's lifecycle to In Review.
  • - Published. Typically used to indicate an asset that is reviewed and accepted by stakeholders and is ready for use by the wider organization.
  • - Obsolete. Typically used to indicate an asset that is no longer in use. For an asset with the lifecycle status as Obsolete, you can only change the status to Published.
Model Format
Optional
The format in which the asset is maintained.
You can select one of the following options:
  • - Docker. Docker allows creation of an operating system bundle with deep learning environment and packaging it within a single file called a docker image.
  • - ONNX. Open Neural Network Exchange (ONNX) is an open format that represents machine learning models, that facilitates interchanging models between various ML frameworks and tools.
  • - Others. If none of the other formats meet your requirements, select this option.
Model Rules
Optional
Define the rules that govern the use of the AI model.
Type of Entry: Formatted text
Output
Optional
Result of the AI model after it is run. Use this field to inform users of what they can expect from the model.
Type of Entry: Formatted text
Source Model Repository
Optional
Repository in which the source of the AI model is maintained.
Type of Entry: Alphanumeric characters, including spaces and special characters.
Operation
Required
Select an operation for the asset.
You can select one of the following options:
  • - Create. You want to create an asset.
  • - Update. You want to modify an existing asset. When you select the Update operation, ensure that you specify a value in the Reference ID column and in the columns that you want to modify. For example, if you want to modify the Lifecycle of an asset, you must enter the Reference ID and Lifecycle values, but you don't need to specify values for any other columns, including the columns that are marked as required.
  • - Delete. You want to delete an existing asset. When you select the Delete operation, ensure that you specify a value in the Reference ID column. You don't need to specify values for any other columns, including the columns that are marked as required. Before you delete an asset, remove or delete all associated relationships and parent-child relationships of the asset.
Stakeholder: Governance Administrator
Optional
User or user group that you want to assign as stakeholders of the asset under the Governance Administrator role. If you enter the name of the user, the user becomes a stakeholder of the asset. If you enter a user group here, all users that belong to the user group become stakeholders of the asset. You can also remove a user or user group as a stakeholder of the asset.
Ensure that the users or user groups that you specify are already assigned the relevant user role. For example, to add a user as a stakeholder with the Governance Owner stakeholder role on an asset, the user must already be assigned the Governance Owner user role from the Administrator console.
Type of Entry: Internal ID, email address of the user, first name followed by the last name with a space in between, or last name followed by comma and then the first name, without any spaces in between.
Note: To retrieve the internal ID of a user account or a user group, go to Administrator. On the respective user or user group page, the URL contains the internal ID.
Example: 23XVYy0NVSdiVYcf4ItJaQ is the internal ID for the following user page URL:
https://usw5.dm-us.informaticacloud.com/cloudUI/products/administer/main/usersAsset/23XVYy0NVSdiVYcf4ItJaQ
To enter multiple values, use a semicolon to separate each value. Do not use a space to separate different values.
If you want to add a user and a user group as the stakeholders of an asset, ensure that you enter the user names and the user group names in separate rows.
To add a stakeholder, you can optionally prefix the term ADD: to the user or user group. To remove a stakeholder, you must mandatorily prefix the term DELETE: to the user or user group.
Examples:
To add users: “23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
To add users using the ADD prefix: “ADD:23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
To remove users using the DELETE prefix: “DELETE:23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
Stakeholder: (Stakeholder role name)
Optional
User or user group that you want to assign as stakeholders of the asset under the (Stakeholder role name) stakeholder role. If you enter the name of the user, the user becomes a stakeholder of the asset. If you enter a user group here, all users that belong to the user group become stakeholders of the asset. You can also remove a user or user group as a stakeholder of the asset.
Ensure that the users or user groups that you specify are already assigned the relevant user role. For example, to add a user as a stakeholder with the Governance Owner stakeholder role on an asset, the user must already be assigned the Governance Owner user role from the Administrator console.
Type of Entry: Internal ID, email address of the user, first name followed by the last name with a space in between, or last name followed by comma and then the first name, without any spaces in between.
Note: To retrieve the internal ID of a user account or a user group, go to Administrator. On the respective user or user group page, the URL contains the internal ID.
Example: 23XVYy0NVSdiVYcf4ItJaQ is the internal ID for the following user page URL:
https://usw5.dm-us.informaticacloud.com/cloudUI/products/administer/main/usersAsset/23XVYy0NVSdiVYcf4ItJaQ
To enter multiple values, use a semicolon to separate each value. Do not use a space to separate different values.
If you want to add a user and a user group as the stakeholders of an asset, ensure that you enter the user names and the user group names in separate rows.
To add a stakeholder, you can optionally prefix the term ADD: to the user or user group. To remove a stakeholder, you must mandatorily prefix the term DELETE: to the user or user group.
Examples:
To add users: “23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
To add users using the ADD prefix: “ADD:23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
To remove users using the DELETE prefix: “DELETE:23XVYy0NVSdiVYcf4ItJaQ; steve@yourorg.com;John Admin;Rhodes,Jane”
The template displays all the stakeholder roles that are relevant for the asset. For more information about stakeholder roles, see the Asset Management help.