17769614. MODEL GENERATION SYSTEM, METHOD, AND PROGRAM simplified abstract (NEC Corporation)

From WikiPatents
Jump to navigation Jump to search

MODEL GENERATION SYSTEM, METHOD, AND PROGRAM

Organization Name

NEC Corporation

Inventor(s)

Makoto Takamoto of Tokyo (JP)

MODEL GENERATION SYSTEM, METHOD, AND PROGRAM - A simplified explanation of the abstract

This abstract first appeared for US patent application 17769614 titled 'MODEL GENERATION SYSTEM, METHOD, AND PROGRAM

Simplified Explanation

The abstract describes a patent application for a system that generates false data using a neural network model and evaluates the likeness of this data to true data using another neural network model. The system then calculates gradient information to update the weights of the discriminator model in order to increase the distance between the output values for true and false data.

  • Data generation unit generates false data based on a neural network model.
  • Discriminator unit evaluates the likeness of given data to true data using another neural network model.
  • Gradient information calculation unit calculates update amounts for the weights of the discriminator model to increase the distance between output values for true and false data.
  • Selection unit chooses false data to be stored in a saved data storage unit.

Potential Applications

This technology could be applied in various fields such as data augmentation, anomaly detection, and adversarial training in machine learning models.

Problems Solved

This technology helps in generating diverse and realistic false data for training machine learning models, improving the performance of these models in various tasks.

Benefits

The system allows for more robust training of machine learning models by providing a larger and more varied dataset, leading to better generalization and performance.

Potential Commercial Applications

One potential commercial application of this technology could be in the development of advanced data augmentation tools for companies working with large datasets in fields such as image recognition, natural language processing, and fraud detection.

Possible Prior Art

Prior art in this field includes various techniques for data augmentation and adversarial training in machine learning, but this specific system combining neural networks for data generation and evaluation may be a novel approach.

Unanswered Questions

How does this technology compare to existing data augmentation methods in terms of performance and efficiency?

This article does not provide a direct comparison with existing data augmentation methods, leaving the reader to wonder about the relative advantages of this technology.

What are the potential limitations or challenges in implementing this system in real-world applications?

The article does not address any potential limitations or challenges that may arise when implementing this system in practical settings, leaving room for uncertainty regarding its feasibility and scalability.


Original Abstract Submitted

A data generation unit generates a first number of false data based on a generation model that is a neural network for generating false data. A discriminator unit derives output values for given data based on a discriminator model that is a neural network for deriving output values indicating true data-likeness and false data-likeness of the given data. A gradient information calculation unit calculates, for each combination of one true data and each of the first number of false data, a distance between the output value for the true data and the output value for the false data, and calculates gradient information that is an update amount for each weight that the discriminator model has, so as to increase the distance by a predetermined amount. A selection unit selects false data to be stored in the saved data storage unit