Resource: BacktestResult
BacktestResult is created to test the performance of a model on a dataset.
| JSON representation |
|---|
{ "name": string, "createTime": string, "updateTime": string, "labels": { string: string, ... }, "state": enum ( |
| Fields | |
|---|---|
name |
Output only. The resource name of the BacktestResult. format: |
createTime |
Output only. The timestamp of creation of this resource. Uses RFC 3339, where generated output will always be Z-normalized and use 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
updateTime |
Output only. The timestamp of the most recent update of this resource. Uses RFC 3339, where generated output will always be Z-normalized and use 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
labels |
Labels An object containing a list of |
state |
Output only. State of the BacktestResult (creating, active, deleting, etc.) |
dataset |
Required. The resource name of the Dataset to backtest on Format: |
model |
Required. The resource name of the Model to use or to backtest. Format: |
endTime |
Required. End_time specifies the latest time from which labels are used and from which data is used to generate features for backtesting. End_time should be no later than the end of the dateRange of the primary dataset. Uses RFC 3339, where generated output will always be Z-normalized and use 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
backtestPeriods |
The number of consecutive months to conduct backtesting for, ending with the last full month prior to the endTime according to the dataset's timezone. |
performanceTarget |
Required. PerformanceTarget gives information on how the test will be evaluated. |
lineOfBusiness |
Output only. The line of business (Retail/Commercial) this backtest is for. Determined by Model, cannot be set by user. |
satisfiesPzi |
Output only. [Output Only] Reserved for future use. |
satisfiesPzs |
Output only. [Output Only] Reserved for future use. |
State
The possible states of a resource.
| Enums | |
|---|---|
STATE_UNSPECIFIED |
State is unspecified, should not occur. |
CREATING |
The resource has not finished being created. |
ACTIVE |
The resource is active/ready to be used. |
UPDATING |
The resource is in the process of being updated. |
DELETING |
The resource is in the process of being deleted. |
PerformanceTarget
PerformanceTarget gives hints on how to evaluate the performance of a model.
| JSON representation |
|---|
{ "partyInvestigationsPerPeriodHint": string, "partyCountExceedingThresholdPerPeriodHint": string } |
| Fields | |
|---|---|
partyInvestigationsPerPeriodHint |
Optional. A hint on the number of parties from this data that will be investigated per period (monthly). This is used to ensure the model is evaluated at relevant investigation volumes. For example, when trying AML AI for the first time, we recommend setting this to the number of parties investigated in an average month, based on alerts from your existing automated alerting system. Either this field or partyCountExceedingThresholdPerPeriodHint should be provided. If both are omitted, or both are provided, it will throw an error. |
partyCountExceedingThresholdPerPeriodHint |
Optional. A hint on the number of parties from this data that should score above the threshold for alerting per period (monthly). This is used to ensure the model is evaluated at a relevant alerting volume and can be used for a quick comparison of backtesting for multiple candidate models. Either this field or partyInvestigationsPerPeriodHint should be provided. If both are omitted, or both are provided, it will throw an error. |
Methods |
|
|---|---|
|
Create a BacktestResult. |
|
Deletes a BacktestResult. |
|
Export governance information for a BacktestResult resource. |
|
Gets a BacktestResult. |
|
List BacktestResults. |
|
Updates the parameters of a single BacktestResult. |