Feature processing recipes for machine learning
First Claim
Patent Images
1. A system, comprising:
- one or more computing devices configured to;
implement a network-based interface for a network-accessible machine learning service of a provider network, the network-based interface configured to receive text representations of different client recipes from a plurality of respective clients of the network-accessible machine learning service, wherein respective client recipes received via the network-based interface specify one or more transformations to be performed on respective data sets used for model training or prediction by the machine learning service, the respective text representations of client recipes comprising one or more of;
(a) a group definitions section indicating one or more groups of variables, wherein individual ones of the one or more groups comprise a plurality of variables on which at least one common transformation operation is to be applied, (b) an assignment section defining one or more intermediate variables, (c) a dependency section indicating respective references to one or more machine learning artifacts stored in a repository, or (d) an output section indicating one or more transformation operations to be applied to at least one entity indicated in the group definitions section, the assignment section, or the dependency section;
validate, in accordance with (a) a set of syntax rules defined by the machine learning service and (b) a set of library function definitions for transformation operation types supported by the machine learning service, a given text representation of a given client recipe received via the network-based interface;
generate, from the given text representation, an executable representation of the given client recipe;
store the executable representation in the repository;
determine that the given client recipe is to be applied to a particular data set;
verify that the particular data set meets a run-time acceptance criterion of the given client recipe; and
execute the executable representation of the given client recipe, using one or more selected provider network resources, to apply a particular transformation operation specified in the given client recipe to the particular data set.
1 Assignment
0 Petitions
Accused Products
Abstract
A first representation of a feature processing recipe is received at a machine learning service. The recipe includes a section in which groups of variables on which common transformations are to be applied are defined, and a section in which a set of transformation operations are specified. The first representation of the recipe is validated based at least in part on a library of function definitions supported by the service, and an executable version of the recipe is generated. In response to a determination that the recipe is to be executed on a particular data set, a set of provider network resources is used to implement a transformation operation indicated in the recipe.
-
Citations
27 Claims
-
1. A system, comprising:
one or more computing devices configured to; implement a network-based interface for a network-accessible machine learning service of a provider network, the network-based interface configured to receive text representations of different client recipes from a plurality of respective clients of the network-accessible machine learning service, wherein respective client recipes received via the network-based interface specify one or more transformations to be performed on respective data sets used for model training or prediction by the machine learning service, the respective text representations of client recipes comprising one or more of;
(a) a group definitions section indicating one or more groups of variables, wherein individual ones of the one or more groups comprise a plurality of variables on which at least one common transformation operation is to be applied, (b) an assignment section defining one or more intermediate variables, (c) a dependency section indicating respective references to one or more machine learning artifacts stored in a repository, or (d) an output section indicating one or more transformation operations to be applied to at least one entity indicated in the group definitions section, the assignment section, or the dependency section;validate, in accordance with (a) a set of syntax rules defined by the machine learning service and (b) a set of library function definitions for transformation operation types supported by the machine learning service, a given text representation of a given client recipe received via the network-based interface; generate, from the given text representation, an executable representation of the given client recipe; store the executable representation in the repository; determine that the given client recipe is to be applied to a particular data set; verify that the particular data set meets a run-time acceptance criterion of the given client recipe; and execute the executable representation of the given client recipe, using one or more selected provider network resources, to apply a particular transformation operation specified in the given client recipe to the particular data set. - View Dependent Claims (2, 3, 4, 5)
-
6. A method, comprising:
performing, by one or more computing devices; implementing a network-based interface for a network-accessible machine learning service of a provider network, the network-based interface configured to receive representations of different client recipes from a plurality of respective clients of the network-accessible machine learning service, wherein respective client recipes received via the network-based interface specify one or more transformations to be performed on respective data sets used for model training or prediction by the machine learning service, the respective representations of client recipes comprising one or more of;
(a) a group definitions section indicating one or more groups of variables, wherein individual ones of the one or more groups comprise a plurality of data set variables on which at least one common transformation operation is to be applied and (b) an output section indicating one or more transformation operations to be applied to at least one entity indicated in one or more of;
(i) the group definitions section or (ii) an input data set;validating, in accordance with at least a set of library function definitions for transformation operation types supported by the machine learning service, a given representation of a given client recipe received via the network-based interface; generating, from the given representation, an executable representation of the given client recipe; determining that the given client recipe is to be applied to a particular data set; verifying that the particular data set meets a run-time acceptance criterion; and executing, using one or more selected provider network resources, the executable representation of the given client recipe to apply a particular transformation operation specified in the given client recipe to the particular data set. - View Dependent Claims (7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22)
-
23. A non-transitory computer-accessible storage medium storing program instructions that when executed on one or more processors:
-
implement a network-based interface for a network-accessible machine learning service of a provider network, the network-based interface configured to receive representations of different client recipes from a plurality of respective clients of the network-accessible machine learning service, wherein respective client recipes received via the network-based interface specify one or more transformations to be performed on respective data sets used for model training or prediction by the machine learning service, the respective representations of client recipes comprising one or more of;
(a) a group definitions section indicating one or more groups of variables, wherein individual ones of the one or more groups comprise a plurality of data set variables on which at least one common transformation operation is to be applied, or (b) an output section indicating one or more transformation operations to be applied to at least one entity indicated in one or more of (i) the group definitions section or (ii) an input data set of the recipe;validate, in accordance with at least a set of library function definitions for transformation operation types supported by the machine learning service, a given representation of a given client recipe received via the network-based interface; generate an executable representation of the given client recipe; and in response to a determination that the given client recipe is to be applied to a particular data set, execute the executable representation of the given client recipe using one or more selected provider network resources to apply a particular transformation operation of the given client recipe to the particular data set. - View Dependent Claims (24, 25, 26, 27)
-
Specification