17995300. METHOD FOR EFFICIENT DISTRIBUTED MACHINE LEARNING HYPERPARAMETER SEARCH simplified abstract (Telefonaktiebolaget LM Ericsson (publ))

From WikiPatents
Jump to navigation Jump to search

METHOD FOR EFFICIENT DISTRIBUTED MACHINE LEARNING HYPERPARAMETER SEARCH

Organization Name

Telefonaktiebolaget LM Ericsson (publ)

Inventor(s)

Hasan Farooq of Santa Clara CA (US)

Julien Forgeat of San Jose CA (US)

Meral Shirazipour of Santa Clara CA (US)

METHOD FOR EFFICIENT DISTRIBUTED MACHINE LEARNING HYPERPARAMETER SEARCH - A simplified explanation of the abstract

This abstract first appeared for US patent application 17995300 titled 'METHOD FOR EFFICIENT DISTRIBUTED MACHINE LEARNING HYPERPARAMETER SEARCH

Simplified Explanation

The abstract describes a method to improve the efficiency of hyperparameter search in a self-organizing network (SON) by using a hyperparameter server and edge devices. Here are the key points:

  • The method involves sending configuration for data feature collection to edge devices in a self-organizing network.
  • The hyperparameter server receives hyperparameter performance data from the edge devices.
  • A shared hyperparameter machine learning model is trained using the global training database, which includes the hyperparameter performance data.
  • The trained model identifies optimal hyperparameters for the edge devices to use.
  • The edge devices receive configuration for data feature collection from the hyperparameter server.
  • The edge devices train their own machine learning models using local training data and the selected hyperparameters.
  • Performance data obtained from the training of the edge machine learning models is sent back to the hyperparameter server.

Potential applications of this technology:

  • Improving the efficiency of hyperparameter search in self-organizing networks.
  • Enhancing the performance of machine learning models in edge devices.
  • Optimizing hyperparameters for various tasks in a self-organizing network.

Problems solved by this technology:

  • Hyperparameter search can be time-consuming and resource-intensive.
  • Finding optimal hyperparameters for edge devices in a self-organizing network can be challenging.
  • The method provides a systematic approach to efficiently search for hyperparameters in a distributed network.

Benefits of this technology:

  • Reduces the time and resources required for hyperparameter search.
  • Improves the performance of machine learning models in edge devices.
  • Enables optimal hyperparameter selection for different tasks in a self-organizing network.


Original Abstract Submitted

A method of a hyperparameter server improves hyper-parameter search efficiency for devices in a self-organizing network (SON) includes sending configuration for data feature collection to at least one edge device in the self-organizing network, receiving hyper-parameter performance data from the at least one edge device, and training a shared hyperparameter machine learning model using a global training database including the hyperparameter performance data to identify optimal hyperparameters for use by the at least one edge device. A further method of an edge device improves hyperparameter search efficiency for devices in a SON includes receiving configuration for data feature collection from a hyperparameter server, training an edge machine learning model using local training data and selected hyperparameters, and sending performance data to the hyperparameter server obtained from the training of the edge machine learning model.

Retrieved from "http://wikipatents.org/index.php?title=17995300._METHOD_FOR_EFFICIENT_DISTRIBUTED_MACHINE_LEARNING_HYPERPARAMETER_SEARCH_simplified_abstract_(Telefonaktiebolaget_LM_Ericsson_(publ))&oldid=25327"