Generation of diverse simulated data
A system and method include reception of a data object template comprising a plurality of fields and a respective value for each of the plurality of fields, determination of at least one master data-dependent field of the plurality of fields, generation of a plurality of data instances comprising the respective value for each of the plurality of fields except for the at least one master data-dependent field, where a value of the at least one master data-dependent field in each of the plurality of data instances is different from the respective value of the at least one master data-dependent field in the data object template, input of each of the plurality of data instances into a machine learning model to determine a likelihood of successful deployment for each of the plurality of data instances, and determination of a second plurality of data instances for deployment based on the determined likelihoods.
1. A system comprising:
a memory storing program code; and
at least one processing unit to execute the program code to cause the system to:
receive a data object template comprising a plurality of fields and a respective value for each of the plurality of fields;
determine at least one master data-dependent field of the plurality of fields;
determine a desired number of data instances;
generate a plurality of data instances based on the desired number of data instances, each of the plurality of data instances comprising the respective value for each of the plurality of fields except for the at least one master data-dependent field, where a value of the at least one master data-dependent field in each of the plurality of data instances is different from the respective value of the at least one master data-dependent field in the data object template;
determine, for each of the plurality of data instances and using a machine learning model, a likelihood of successful deployment; and
determine a second plurality of the plurality of data instances for deployment by determining ones of the plurality of data instances which are associated with a likelihood based on the determined likelihoods of successful deployment greater than a threshold.
2. A system according to claim 1 , wherein the plurality of data instances are generated based on a data dictionary of values associated with the at least one master data-dependent field.
3. A system according to claim 2 , wherein the template identifies the at least one master data-dependent field.
4. A system according to claim 1 , wherein the template identifies the at least one master data-dependent field.
5. A system according to claim 1 , the at least one processing unit to execute the program code to cause the system to:
determine a desired number of data instances; and
determine to generate a multiple of the desired number of data instances.
6. A system according to claim 5 , wherein determination of the second plurality of the plurality of data instances comprises determination of the desired number of data instances based on the determined likelihoods of successful deployment.
7. A system according to claim 1 , wherein determination of the second plurality of the plurality of data instances comprises determination of ones of the plurality of data instances which are associated with a likelihood of successful deployment greater than a threshold.
8. A method comprising:
receiving a data object template comprising a plurality of fields and a respective value for each of the plurality of fields;
determining at least one master data-dependent field of the plurality of fields;
determining a desired number of data instances;
generating a plurality of data instances, based on the desired number of data instances, each of the plurality of data instances comprising the respective value for each of the plurality of fields except for the at least one master data-dependent field, where a value of the at least one master data-dependent field in each of the plurality of data instances is determined based on a data dictionary;
inputting each of the plurality of data instances into a machine learning model to | determine a likelihood of successful deployment for each of the plurality of data instances; and determining a second plurality of the plurality of data instances for deployment based on the determined likelihoods based on the determined likelihoods of successful deployment greater than a threshold; and
deploying the second plurality of the plurality of data instances to a target system.
9. A method according to claim 8 , wherein the data dictionary comprises values associated with the at least one master data-dependent field.
10. A method according to claim 9 , wherein the at least one master data-dependent field is determined based on the template.
11. A method according to claim 8 , wherein the at least one master data-dependent field is determined based on the template.
12. A method according to claim 8 , further comprising:
determining a desired number of data instances; and
determining to generate a multiple of the desired number of data instances.
13. A method according to claim 12 , wherein determining the second plurality of the plurality of data instances comprises determining the desired number of data instances based on the determined likelihoods of successful deployment.
14. A method according to claim 8 , wherein determining the second plurality of the plurality of data instances comprises determining ones of the plurality of data instances which are associated with a likelihood of successful deployment greater than a threshold.
15. A non-transitory computer-readable recording medium storing program code, the program code executable by at least one processing unit of a computing system to:
receive a data object template comprising a plurality of fields and a respective value for each of the plurality of fields;
determine at least one master data-dependent field of the plurality of fields;
determine a desired number of data instances;
generate a plurality of data instances, based on the desired number of data instances, each of the plurality of data instances comprising the respective value for each of the plurality of fields except for the at least one master data-dependent field, where a value of the at least one master data-dependent field in each of the plurality of data instances is different from the respective value of the at least one master data-dependent field in the data object template;
input each of the plurality of data instances into a machine learning model to determine a likelihood of successful deployment for each of the plurality of data instances; and
determine a second plurality of the plurality of data instances for deployment based on the determined likelihoods based on the determined likelihoods of successful deployment greater than a threshold; and
deploy the second plurality of the plurality of data instances to a target system.
16. A medium according to claim 15 , wherein the plurality of data instances are generated based on a data dictionary of values associated with the at least one master data-dependent field.
17. A medium according to claim 16 , wherein the at least one master data-dependent field is determined based on the template.
18. A medium according to claim 15 , wherein the at least one master data-dependent field is determined based on the template.
19. A medium according to claim 15 , the program code executable by at least one processing unit of a computing system to:
determine a desired number of data instances; and
determine to generate a multiple of the desired number of data instances, wherein determination of the second plurality of the plurality of data instances comprises determination of the desired number of data instances based on the determined likelihoods of successful deployment.
20. A medium according to claim 15 , wherein determination of the second plurality of the plurality of data instances comprises determination of ones of the plurality of data instances which are associated with a likelihood of successful deployment greater than a threshold.