17849506. GENERATING TRAINING DATA USING SAMPLED VALUES simplified abstract (Microsoft Technology Licensing, LLC)

From WikiPatents
Jump to navigation Jump to search

GENERATING TRAINING DATA USING SAMPLED VALUES

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

Gopiram Roshan Lal of Sunnyvale CA (US)

Girish Kathalagiri of San Jose CA (US)

Alice Hing-Yee Leung of Sunnyvale CA (US)

Daqian Sun of Ithaca NY (US)

Aman Grover of San Carlos CA (US)

GENERATING TRAINING DATA USING SAMPLED VALUES - A simplified explanation of the abstract

This abstract first appeared for US patent application 17849506 titled 'GENERATING TRAINING DATA USING SAMPLED VALUES

Simplified Explanation

The abstract describes a method, system, and apparatus for generating a trained neural network prediction model. Here are the key points:

  • The method involves determining a set of data that includes numerical ranges associated with an embedding and an attribute.
  • The numerical range is sampled to obtain a sample value, which is also associated with the embedding and the attribute.
  • A set of sample value training data is generated, including the sample value, the associated embedding, and the associated attribute.
  • A trained neural network prediction model is created by applying a prediction model to the set of sample value training data.
  • The trained neural network prediction model is then used to determine an output based on a set of input data.
  • The output is a predicted range of values, calculated using an output mean and an output standard deviation.

Potential Applications:

  • Predictive analytics in various fields such as finance, healthcare, and manufacturing.
  • Forecasting future trends or outcomes based on input data and associated attributes.
  • Optimizing decision-making processes by providing predicted ranges of values.

Problems Solved:

  • Overcoming the limitations of traditional prediction models by utilizing neural networks.
  • Providing a more accurate and reliable prediction model by incorporating sample value training data.
  • Addressing the need for predicting ranges of values rather than single point estimates.

Benefits:

  • Improved accuracy in predicting ranges of values, allowing for more informed decision-making.
  • Increased flexibility in handling complex data sets with multiple numerical ranges and attributes.
  • Enhanced efficiency in generating trained neural network prediction models for various applications.


Original Abstract Submitted

Methods, systems, and apparatuses include determining a set of data. The set of data includes multiple numerical ranges associated with an embedding and an attribute. The numerical range is sampled to obtain a sample value which is also associated with the embedding and the attribute. A set of sample value training data is generated, the set including the sample value, the associated embedding, and the associated attribute. A trained neural network prediction model is generated by applying a prediction model to the set of sample value training data. A set of input data is applied to the trained neural network prediction model. An output is determined by the trained neural network prediction model based on the set of input data. The output is a predicted range of values based on an output mean and an output standard deviation.

Retrieved from "http://wikipatents.org/index.php?title=17849506._GENERATING_TRAINING_DATA_USING_SAMPLED_VALUES_simplified_abstract_(Microsoft_Technology_Licensing,_LLC)&oldid=17087"