Synchronized distributed processing in a communications network

1. A method for implementing distributed processing in edge nodes, the method comprising:

for each respective edge node of a plurality of edge nodes:

receiving, at the respective edge node, data from a client of the respective edge node;

storing the received data to memory of the respective edge node;

generating, at the respective edge node, a prediction of a respective activity based on the received data;

determining, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model, wherein the machine learning model is generated by:

identifying a subset of the plurality of edge nodes comprising common attributes that are common with respect to attributes of the respective edge node;

retrieving respective models for each respective edge node of the subset;

weighting aspects of each respective model;

ranking each respective aspect of each respective model based on the weighting; and

initializing the machine learning model for the respective edge node by generating a machine learning model comprising aspects selected based on the ranking;

feeding the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, deleting the received data from the memory of the respective edge node; and

in response to determining that the prediction is valid, activating, at the respective edge node, a function.

2. The method of claim 1 , wherein the machine learning model comprises a reinforced learning model.

3. A method for implementing distributed processing in edge nodes, the method comprising:

for each respective edge node of a plurality of edge nodes:

receiving, at the respective edge node, data from a client of the respective edge node;

storing the received data to memory of the respective edge node;

generating, at the respective edge node, a prediction of a respective activity based on the received data;

determining, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model;

feeding the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, deleting the received data from the memory of the respective edge node;

in response to determining that the prediction is valid, activating, at the respective edge node, a function;

in response to determining that the prediction is valid, determining whether the respective activity is novel;

in response to determining that the respective activity is novel:

identifying a subset of nodes that correlate to the respective node; and

propagating the determination of validity to the subset of nodes.

4. The method of claim 3 , wherein determining whether the respective activity is novel further comprises determining that the respective activity relates to cyber security.

5. The method of claim 3 , wherein propagating the determination of validity to the subset of nodes comprises:

identifying confidential information in data to be propagated; and

obscuring the confidential data from the data to be propagated.

6. A system comprising for implementing distributed processing in edge nodes, the system comprising:

a non-transitory computer-readable medium with instructions encoded thereon for implemented distributed processing in edge nodes; and

a processor configured, based on execution of the instructions, to, for each respective edge node of a plurality of edge nodes:

receive, at the respective edge node, data from a client of the respective edge node;

store the received data to memory of the respective edge node;

generate, at the respective edge node, a prediction of a respective activity based on the received data;

determine, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model, wherein the machine learning model is generated by:

identifying a subset of the plurality of edge nodes comprising common attributes that are common with respect to attributes of the respective edge node;

retrieving respective models for each respective edge node of the subset;

weighting aspects of each respective model;

ranking each respective aspect of each respective model based on the weighting; and

initializing the machine learning model for the respective edge node by generating a machine learning model comprising aspects selected based on the ranking;

feed the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, delete the received data from the memory of the respective edge node; and

in response to determining that the prediction is valid, activate, at the respective edge node, a function.

7. The system of claim 6 , wherein the machine learning model comprises a reinforced learning model.

8. A system comprising for implementing distributed processing in edge nodes, the system comprising:

a non-transitory computer-readable medium with instructions encoded thereon for implemented distributed processing in edge nodes; and

a processor configured, based on execution of the instructions, to, for each respective edge node of a plurality of edge nodes:

receive, at the respective edge node, data from a client of the respective edge node;

store the received data to memory of the respective edge node;

generate, at the respective edge node, a prediction of a respective activity based on the received data;

determine, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model;

feed the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, delete the received data from the memory of the respective edge node;

in response to determining that the prediction is valid, activating, at the respective edge node, a function;

in response to determining that the prediction is valid, determine whether the respective activity is novel;

in response to determining that the respective activity is novel:

identify a subset of nodes that correlate to the respective node; and

propagate the determination of validity to the subset of nodes.

9. The system of claim 8 , wherein the processor is further configured, when determining whether the respective activity is novel further, to determine that the respective activity relates to cyber security.

10. The system of claim 8 , wherein the processor is further configured, when propagating the determination of validity to the subset of nodes, to:

identify confidential information in data to be propagated; and

obscure the confidential data from the data to be propagated.

11. A non-transitory computer-readable medium with instructions encoded thereon for implementing distributed processing in edge nodes, the instructions comprising instructions to:

for each respective edge node of a plurality of edge nodes:

receive, at the respective edge node, data from a client of the respective edge node;

store the received data to memory of the respective edge node;

generate, at the respective edge node, a prediction of a respective activity based on the data;

determine, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model, wherein the machine learning model is generated by:

identifying a subset of the plurality of edge nodes comprising common attributes that are common with respect to attributes of the respective edge node;

retrieving respective models for each respective edge node of the subset;

weighting aspects of each respective model;

ranking each respective aspect of each respective model based on the weighting; and

initializing the machine learning model for the respective edge node by generating a machine learning model comprising aspects selected based on the ranking;

feed the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, delete the received data from the memory of the respective edge node; and

in response to determining that the prediction is valid, activate, at the respective edge node, a function.

12. The non-transitory computer-readable medium of claim 11 , wherein the machine learning model comprises a reinforced learning model.

13. A non-transitory computer-readable medium with instructions encoded thereon for implementing distributed processing in edge nodes, the instructions comprising instructions to:

for each respective edge node of a plurality of edge nodes:

receive, at the respective edge node, data from a client of the respective edge node;

store the received data to memory of the respective edge node;

generate, at the respective edge node, a prediction of a respective activity based on the data;

determine, at the respective edge node, whether the prediction is valid by feeding the prediction into a machine learning model and receiving a validation response from the machine learning model;

feed the received data as training data to the machine learning model;

responsive to having trained the machine learning model using the received data, delete the received data from the memory of the respective edge node;

in response to determining that the prediction is valid, activate, at the respective node, a function;

in response to determining that the prediction is valid, determine whether the respective activity is novel;

in response to determining that the respective activity is novel:

identify a subset of nodes that correlate to the respective node; and

propagate the determination of validity to the subset of nodes.

14. The non-transitory computer-readable medium of claim 11 , wherein the instructions to propagate the determination of validity to the subset of nodes further comprise instructions to:

identify confidential information in data to be propagated; and

obscure the confidential data from the data to be propagated.