Staged training of neural networks for improved time series prediction performance

• US 10,650,045 B2
• Filed: 08/30/2019
• Issued: 05/12/2020
• Est. Priority Date: 02/05/2016
• Status: Active Grant

• Current Assignee: SAS Institute Incorporated
• Original Assignee: SAS Institute Incorporated
• Inventors: Bequet, Henry Gabriel Victor, Rioux, Jacques, Izquierdo, John Alejandro, Chen, Huina, Du, Juan
• Primary Examiner(s): Patel, Hiren P
• Application Number:
  US16/556,573
• Publication Number:
  US 20190384790A1
• Time in Patent Office: 256 Days
• Field of Search: None
• CPC Class Codes: G06F 16/9014 hash tables
• CPC Class Codes: G06F 16/903 Querying for retrieval from the web G06F16/953
• CPC Class Codes: G06F 16/90344 by using string matching techniques
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/063 using electronic means
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent
• CPC Class Codes: H04L 41/16 using machine learning or artificial intelligence
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46

Abstract

An apparatus includes a processor to: train a first neural network of a chain to generate first configuration data including first trained parameters, wherein the chain performs an analytical function generating a set of output values from a set of input values, each neural network has inputs to receive the set of input values and outputs to output a portion of the set of output values, and the neural networks are ordered from the first at the head to a last neural network at the tail, and are interconnected so that each neural network additionally receives the outputs of a preceding neural network; train, using the first configuration data, a next neural network in the chain ordering to generate next configuration data including next trained parameters; and use at least the first and next configuration data and data indicating the interconnections to instantiate the chain to perform the analytical function.

Claims

1. An apparatus comprising a processor and a storage to store instructions that, when executed by the processor, cause the processor to perform operations comprising:

train, using initial neural network configuration data comprising neural network hyperparameters, a first neural network of a chain of neural networks to generate first neural network configuration data comprising the hyperparameters and first trained parameters learned by the first neural network, wherein;

the chain is to perform an analytical function to generate a set of output data values from a set of input data values;

the analytical function comprises the generation of a time series prediction that covers a selected full range of time;

the chain comprises a set of multiple neural networks that includes at least the first neural network and a last neural network;

the set of neural networks is ordered to form the chain starting with the first neural network at a head of the chain and ending with the last neural network at a tail of the chain;

each neural network in the chain comprises external inputs to receive the set of input data values;

each neural network in the chain comprises outputs at which the neural network outputs a portion of the set of output data values from the set input data values during operation of the chain to perform the analytical function; and

the set of neural networks is interconnected within the chain such that each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of a preceding neural network in the ordering of neural networks within the chain as additional inputs;

train, using the first neural network configuration data, a next neural network in the ordering of neural networks within the chain to generate a next neural network configuration data comprising the hyperparameters and next trained parameters learned by the next neural network;

use at least the first neural network configuration data, the next neural network configuration data, and additional data comprising an indication of interconnections among the neural networks within the chain to instantiate the chain of neural networks; and

operate the chain of neural networks to perform the analytical function.

2. The apparatus of claim 1, wherein:

the full range of time is divided into multiple subranges of time;

the outputs of each neural network of the chain provides the output data values of a portion of the time series prediction that covers one of the multiple subranges of time; and

the subranges are temporally ordered to follow the order of neural networks within the chain wherein;

the outputs of the first neural network at the head of the chain cover the temporally earliest subrange of time; and

the outputs of the last neural network at the tail of the chain cover the temporally latest subrange of time.

3. The apparatus of claim 1, wherein the processor is caused to perform operations comprising:

train each neural network in the chain using neural network training data comprising sets of input values and corresponding sets of output values generated through performances of the analytical function using non-neuromorphic processing; and

train each neural network in the chain via backpropagation wherein;

sets of input values of the neural network training data are provided to the external inputs;

portions of corresponding sets of output values of the neural network training data are provided to the outputs;

the portions of the corresponding sets of output values of the neural network training data that are provided to the outputs are selected to correspond to the portion of a set of output values to be output by the neural network during operation of the chain to perform the analytical function;

for each neural network in the chain other than the first neural network at the head of the chain, the preceding neural network is operated to perform a portion of the analytical function, and the outputs of the preceding neural network are provided to additional inputs; and

a null input value is provided to each one of the additional inputs that does not receive an output of another neural network of the set of neural networks.

4. The apparatus of claim 3, wherein the processor is caused to perform operations comprising:

analyze the neural network training data to identify a portion of the output data values across the sets of output values of the neural network training data that shows a relatively high degree of correlation; and

derive a manner of dividing the outputs of the chain into multiple portions that are each output by one of the of the neural networks in chain based, at least in part, on the identified portion that shows the relatively high degree of correlation.

5. The apparatus of claim 1, wherein:

the hyperparameters specify at least a quantity of artificial neurons within each neural network of the set of neural networks and a quantity of layers of artificial neurons within each neural network of the set of neural networks;

the quantity of layers includes an input layer of artificial neurons connected to the external inputs and the additional inputs;

the quantity of layers includes an output layer of artificial neurons connected to the outputs; and

each neural network of the set of neural networks comprises the same quantities of external inputs, additional inputs and outputs.

6. The apparatus of claim 5, wherein the chain comprises a type of chain selected from a group consisting of:

a single-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of the immediately preceding neural network in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable each neural network, except the first neural network at the head of the chain, to receive all of the outputs of the immediately preceding neural network at its additional inputs; and

a multi-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of all of the preceding neural networks in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable the last neural network to receive all of the outputs of all of the other neural networks in the chain.

7. The apparatus of claim 1, wherein:

the trained parameters of the first neural network configuration data comprise weights and biases that represent what was learned by the first neural network during training; and

the processor is caused to train the set of neural networks sequentially in an order that follows the ordering of neural networks in the chain from the first neural network at the head of the chain to the last neural network at the tail of the chain.

8. The apparatus of claim 1, comprising a plurality of neuromorphic devices communicatively coupled to the processor, wherein the processor is caused to perform operations comprising:

derive the portion of the set of output values that are output by each neural network from the set of input data values during operation of the chain to perform the analytical function based on at least one of;

a quantity of artificial neurons within each neuromorphic device of the plurality of neuromorphic devices;

a maximum quantity of layers that each neuromorphic device of the plurality of neuromorphic devices is able to support;

a maximum quantity of inputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

or

a maximum quantity of outputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

provide the first neural network configuration data to at least a first neuromorphic device of the plurality of neuromorphic devices to instantiate the first neural network;

provide the next neural network configuration data to at least a second neuromorphic device of the plurality of neuromorphic devices to instantiate the second neural network; and

provide a last neural network configuration data to at least a third neuromorphic device of the plurality of neuromorphic devices to instantiate the last neural network.

9. The apparatus of claim 1, wherein the processor is caused to perform operations comprising:

perform a training job flow at least partially within a training federated area to perform the training of each neural network in the chain of neural networks, wherein;

the training federated area is one of multiple federated areas within a hierarchy of federated areas; and

the multiple federated areas are maintained within at least one storage device to store job flow definitions for the training job flow and a testing job flow, task routines executable to perform tasks of each job flow, and data objects for use as inputs to job flows and generated as outputs of performances of job flows;

monitor the training of each neural network in the chain to determine whether a condition is met to transfer a set of instances of neural network configuration data generated by the training to the testing federated area, wherein;

the set of instances of neural network configuration data comprises the first neural network configuration data and the next neural network configuration data; and

in response to the condition being met, the processor is caused to perform operations comprising;

store the set of instances of neural network configuration data as a single data object within the training federated area; and

transfer a copy of the single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the chain to determine whether the chain performs the analytical function with a degree of accuracy that meets at least a predetermined threshold of accuracy.

10. The apparatus of claim 9, wherein the processor is caused to perform operations comprising:

monitor for a transfer of an object to the training federated area from the testing federated area indicative of failure of the chain to perform the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy; and

in response to the transfer of the object indicative of the failure from the testing federated area to the training federated area, the processor is caused to perform operations comprising;

derive a new chain of neural networks to perform the analytical function by changing at least one of;

a quantity of neural networks within the set of neural networks;

the division of the set of output values into portions to be output by each neural network of the set of neural networks from the set of input values during operation of the new chain to perform the analytical function;

or

the interconnections among the neural networks within the new chain;

train each of the neural networks in the new chain to generate a new set of instances of neural network configuration data;

monitor the training of each neural network in the new chain to determine whether the condition is met to transfer the new set of instances of neural network configuration data to the testing federated area; and

in response to the condition being met, the processor is caused to perform operations comprising;

store the new set of instances of neural network configuration data as a new single data object within the training federated area; and

transfer a copy of the new single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the new chain to determine whether the new chain performs the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy.

11. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, the computer-program product including instructions operable to cause a processor to perform operations comprising:

train, using initial neural network configuration data comprising neural network hyperparameters, a first neural network of a chain of neural networks to generate first neural network configuration data comprising the hyperparameters and first trained parameters learned by the first neural network, wherein;

the chain is to perform an analytical function to generate a set of output data values from a set of input data values;

the analytical function comprises the generation of a time series prediction that covers a selected full range of time;

the chain comprises a set of multiple neural networks that includes at least the first neural network and a last neural network;

the set of neural networks is ordered to form the chain starting with the first neural network at a head of the chain and ending with the last neural network at a tail of the chain;

each neural network in the chain comprises external inputs to receive the set of input data values;

each neural network in the chain comprises outputs at which the neural network outputs a portion of the set of output data values from the set input data values during operation of the chain to perform the analytical function; and

the set of neural networks is interconnected within the chain such that each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of a preceding neural network in the ordering of neural networks within the chain as additional inputs;

train, using the first neural network configuration data, a next neural network in the ordering of neural networks within the chain to generate a next neural network configuration data comprising the hyperparameters and next trained parameters learned by the next neural network;

use at least the first neural network configuration data, the next neural network configuration data, and additional data comprising an indication of interconnections among the neural networks within the chain to instantiate the chain of neural networks; and

operate the chain of neural networks to perform the analytical function.

12. The computer-program product of claim 11, wherein:

the full range of time is divided into multiple subranges of time;

the outputs of each neural network of the chain provides the output data values of a portion of the time series prediction that covers one of the multiple subranges of time; and

the subranges are temporally ordered to follow the order of neural networks within the chain wherein;

the outputs of the first neural network at the head of the chain cover the temporally earliest subrange of time; and

the outputs of the last neural network at the tail of the chain cover the temporally latest subrange of time.

13. The computer-program product of claim 11, wherein the processor is caused to perform operations comprising:

train each neural network in the chain using neural network training data comprising sets of input values and corresponding sets of output values generated through performances of the analytical function using non-neuromorphic processing; and

train each neural network in the chain via backpropagation wherein;

sets of input values of the neural network training data are provided to the external inputs;

portions of corresponding sets of output values of the neural network training data are provided to the outputs;

the portions of the corresponding sets of output values of the neural network training data that are provided to the outputs are selected to correspond to the portion of a set of output values to be output by the neural network during operation of the chain to perform the analytical function;

for each neural network in the chain other than the first neural network at the head of the chain, the preceding neural network is operated to perform a portion of the analytical function, and the outputs of the preceding neural network are provided to additional inputs; and

a null input value is provided to each one of the additional inputs that does not receive an output of another neural network of the set of neural networks.

14. The computer-program product of claim 13, wherein the processor is caused to perform operations comprising:

analyze the neural network training data to identify a portion of the output data values across the sets of output values of the neural network training data that shows a relatively high degree of correlation; and

derive a manner of dividing the outputs of the chain into multiple portions that are each output by one of the of the neural networks in chain based, at least in part, on the identified portion that shows the relatively high degree of correlation.

15. The computer-program product of claim 11, wherein:

the hyperparameters specify at least a quantity of artificial neurons within each neural network of the set of neural networks and a quantity of layers of artificial neurons within each neural network of the set of neural networks;

the quantity of layers includes an input layer of artificial neurons connected to the external inputs and the additional inputs;

the quantity of layers includes an output layer of artificial neurons connected to the outputs; and

each neural network of the set of neural networks comprises the same quantities of external inputs, additional inputs and outputs.

16. The computer-program product of claim 15, wherein the chain comprises a type of chain selected from a group consisting of:

a single-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of the immediately preceding neural network in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable each neural network, except the first neural network at the head of the chain, to receive all of the outputs of the immediately preceding neural network at its additional inputs; and

a multi-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of all of the preceding neural networks in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable the last neural network to receive all of the outputs of all of the other neural networks in the chain.

17. The computer-program product of claim 11, wherein:

the trained parameters of the first neural network configuration data comprise weights and biases that represent what was learned by the first neural network during training; and

the processor is caused to train the set of neural networks sequentially in an order that follows the ordering of neural networks in the chain from the first neural network at the head of the chain to the last neural network at the tail of the chain.

18. The computer-program product of claim 11, wherein the processor is caused to perform operations comprising:

derive the portion of the set of output values that are output by each neural network from the set of input data values during operation of the chain to perform the analytical function based on at least one of;

a quantity of artificial neurons within each neuromorphic device of a plurality of neuromorphic devices communicatively coupled to the processor;

a maximum quantity of layers that each neuromorphic device of the plurality of neuromorphic devices is able to support;

a maximum quantity of inputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

or

a maximum quantity of outputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

provide the first neural network configuration data to at least a first neuromorphic device of the plurality of neuromorphic devices to instantiate the first neural network;

provide the next neural network configuration data to at least a second neuromorphic device of the plurality of neuromorphic devices to instantiate the second neural network; and

provide a last neural network configuration data to at least a third neuromorphic device of the plurality of neuromorphic devices to instantiate the last neural network.

19. The computer-program product of claim 11, wherein the processor is caused to perform operations comprising:

perform a training job flow at least partially within a training federated area to perform the training of each neural network in the chain of neural networks, wherein;

the training federated area is one of multiple federated areas within a hierarchy of federated areas; and

the multiple federated areas are maintained within at least one storage device to store job flow definitions for the training job flow and a testing job flow, task routines executable to perform tasks of each job flow, and data objects for use as inputs to job flows and generated as outputs of performances of job flows;

monitor the training of each neural network in the chain to determine whether a condition is met to transfer a set of instances of neural network configuration data generated by the training to the testing federated area, wherein;

the set of instances of neural network configuration data comprises the first neural network configuration data and the next neural network configuration data; and

in response to the condition being met, the processor is caused to perform operations comprising;

store the set of instances of neural network configuration data as a single data object within the training federated area; and

transfer a copy of the single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the chain to determine whether the chain performs the analytical function with a degree of accuracy that meets at least a predetermined threshold of accuracy.

20. The computer-program product of claim 19, wherein the processor is caused to perform operations comprising:

monitor for a transfer of an object to the training federated area from the testing federated area indicative of failure of the chain to perform the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy; and

in response to the transfer of the object indicative of the failure from the testing federated area to the training federated area, the processor is caused to perform operations comprising;

derive a new chain of neural networks to perform the analytical function by changing at least one of;

a quantity of neural networks within the set of neural networks;

the division of the set of output values into portions to be output by each neural network of the set of neural networks from the set of input values during operation of the new chain to perform the analytical function;

or

the interconnections among the neural networks within the new chain;

train each of the neural networks in the new chain to generate a new set of instances of neural network configuration data;

monitor the training of each neural network in the new chain to determine whether the condition is met to transfer the new set of instances of neural network configuration data to the testing federated area; and

in response to the condition being met, the processor is caused to perform operations comprising;

store the new set of instances of neural network configuration data as a new single data object within the training federated area; and

transfer a copy of the new single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the new chain to determine whether the new chain performs the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy.

21. A computer-implemented method comprising:

training, by a processor, and using initial neural network configuration data comprising neural network hyperparameters, a first neural network of a chain of neural networks to generate first neural network configuration data comprising the hyperparameters and first trained parameters learned by the first neural network, wherein;

the chain is to perform an analytical function to generate a set of output data values from a set of input data values;

the analytical function comprises the generation of a time series prediction that covers a selected full range of time;

the chain comprises a set of multiple neural networks that includes at least the first neural network and a last neural network;

the set of neural networks is ordered to form the chain starting with the first neural network at a head of the chain and ending with the last neural network at a tail of the chain;

each neural network in the chain comprises external inputs to receive the set of input data values;

each neural network in the chain comprises outputs at which the neural network outputs a portion of the set of output data values from the set input data values during operation of the chain to perform the analytical function; and

the set of neural networks is interconnected within the chain such that each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of a preceding neural network in the ordering of neural networks within the chain as additional inputs;

training, by the processor, and using the first neural network configuration data, a next neural network in the ordering of neural networks within the chain to generate a next neural network configuration data comprising the hyperparameters and next trained parameters learned by the next neural network;

using at least the first neural network configuration data, the next neural network configuration data, and additional data comprising an indication of interconnections among the neural networks within the chain to instantiate the chain of neural networks; and

operating the chain of neural networks to perform the analytical function.

22. The computer-implemented method of claim 21, wherein:

the full range of time is divided into multiple subranges of time;

the outputs of each neural network of the chain provides the output data values of a portion of the time series prediction that covers one of the multiple subranges of time; and

the subranges are temporally ordered to follow the order of neural networks within the chain wherein;

the outputs of the first neural network at the head of the chain cover the temporally earliest subrange of time; and

the outputs of the last neural network at the tail of the chain cover the temporally latest subrange of time.

23. The computer-implemented method of claim 21, comprising:

training, by the processor, each neural network in the chain using neural network training data comprising sets of input values and corresponding sets of output values generated through performances of the analytical function using non-neuromorphic processing; and

training, by the processor, each neural network in the chain via backpropagation wherein;

sets of input values of the neural network training data are provided to the external inputs;

portions of corresponding sets of output values of the neural network training data are provided to the outputs;

the portions of the corresponding sets of output values of the neural network training data that are provided to the outputs are selected to correspond to the portion of a set of output values to be output by the neural network during operation of the chain to perform the analytical function;

for each neural network in the chain other than the first neural network at the head of the chain, the preceding neural network is operated to perform a portion of the analytical function, and the outputs of the preceding neural network are provided to additional inputs; and

a null input value is provided to each one of the additional inputs that does not receive an output of another neural network of the set of neural networks.

24. The computer-implemented method of claim 23, comprising:

analyzing, by the processor, the neural network training data to identify a portion of the output data values across the sets of output values of the neural network training data that shows a relatively high degree of correlation; and

deriving, by the processor, a manner of dividing the outputs of the chain into multiple portions that are each output by one of the of the neural networks in chain based, at least in part, on the identified portion that shows the relatively high degree of correlation.

25. The computer-implemented method of claim 21, wherein:

the hyperparameters specify at least a quantity of artificial neurons within each neural network of the set of neural networks and a quantity of layers of artificial neurons within each neural network of the set of neural networks;

the quantity of layers includes an input layer of artificial neurons connected to the external inputs and the additional inputs;

the quantity of layers includes an output layer of artificial neurons connected to the outputs; and

each neural network of the set of neural networks comprises the same quantities of external inputs, additional inputs and outputs.

26. The computer-implemented method of claim 25, wherein the chain comprises a type of chain selected from a group consisting of:

a single-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of the immediately preceding neural network in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable each neural network, except the first neural network at the head of the chain, to receive all of the outputs of the immediately preceding neural network at its additional inputs; and

a multi-link chain, wherein;

each neural network in the chain, except the first neural network at the head of the chain, receives the outputs of all of the preceding neural networks in the ordering of neural networks within the chain; and

the quantity of additional inputs is selected to enable the last neural network to receive all of the outputs of all of the other neural networks in the chain.

27. The computer-implemented method of claim 21, wherein:

the trained parameters of the first neural network configuration data comprise weights and biases that represent what was learned by the first neural network during training; and

the method comprises training, by the processor, the set of neural networks sequentially in an order that follows the ordering of neural networks in the chain from the first neural network at the head of the chain to the last neural network at the tail of the chain.

28. The computer-implemented method of claim 21, comprising:

deriving, by the processor, the portion of the set of output values that are output by each neural network from the set of input data values during operation of the chain to perform the analytical function based on at least one of;

a quantity of artificial neurons within each neuromorphic device of a plurality of neuromorphic devices communicatively coupled to the processor;

a maximum quantity of layers that each neuromorphic device of the plurality of neuromorphic devices is able to support;

a maximum quantity of inputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

or

a maximum quantity of outputs that each neuromorphic device of the plurality of neuromorphic devices is able to support;

providing, by the processor, the first neural network configuration data to at least a first neuromorphic device of the plurality of neuromorphic devices to instantiate the first neural network;

providing, by the processor, the next neural network configuration data to at least a second neuromorphic device of the plurality of neuromorphic devices to instantiate the second neural network; and

providing, by the processor, a last neural network configuration data to at least a third neuromorphic device of the plurality of neuromorphic devices to instantiate the last neural network.

29. The computer-implemented method of claim 21, comprising:

performing, by the processor, a training job flow at least partially within a training federated area to perform the training of each neural network in the chain of neural networks;

the training federated area is one of multiple federated areas within a hierarchy of federated areas; and

the multiple federated areas are maintained within at least one storage device to store job flow definitions for the training job flow and a testing job flow, task routines executable to perform tasks of each job flow, and data objects for use as inputs to job flows and generated as outputs of performances of job flows;

monitoring, by the processor, the training of each neural network in the chain to determine whether a condition is met to transfer a set of instances of neural network configuration data generated by the training to the testing federated area, wherein;

the set of instances of neural network configuration data comprises the first neural network configuration data and the next neural network configuration data; and

in response to the condition being met, performing operations comprising;

storing the set of instances of neural network configuration data as a single data object within the training federated area; and

transferring, by the processor, a copy of the single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the chain to determine whether the chain performs the analytical function with a degree of accuracy that meets at least a predetermined threshold of accuracy.

30. The computer-implemented method of claim 29, comprising:

monitoring, by the processor, for a transfer of an object to the training federated area from the testing federated area indicative of failure of the chain to perform the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy; and

in response to the transfer of the object indicative of the failure from the testing federated area to the training federated area, performing operations comprising;

deriving, by the processor, a new chain of neural networks to perform the analytical function by changing at least one of;

a quantity of neural networks within the set of neural networks;

the division of the set of output values into portions to be output by each neural network of the set of neural networks from the set of input values during operation of the new chain to perform the analytical function;

or

the interconnections among the neural networks within the new chain;

training, by the processor, each of the neural networks in the new chain to generate a new set of instances of neural network configuration data;

monitoring, by the processor, the training of each neural network in the new chain to determine whether the condition is met to transfer the new set of instances of neural network configuration data to the testing federated area; and

in response to the condition being met, performing operations comprising;

storing the new set of instances of neural network configuration data as a new single data object within the training federated area; and

transferring, by the processor, a copy of the new single data object to the testing federated area to enable the testing job flow to be at least partially performed within the testing federated area to test the new chain to determine whether the new chain performs the analytical function with a degree of accuracy that meets at least the predetermined threshold of accuracy.