View source for 20240037408. METHOD AND APPARATUS FOR MODEL TRAINING AND DATA ENHANCEMENT, ELECTRONIC DEVICE AND STORAGE MEDIUM simplified abstract (JINGDONG CITY (BEIJING) DIGITS TECHNOLOGY CO., LTD.)

=METHOD AND APPARATUS FOR MODEL TRAINING AND DATA ENHANCEMENT, ELECTRONIC DEVICE AND STORAGE MEDIUM=


==Organization Name==

[[:Category:JINGDONG CITY (BEIJING) DIGITS TECHNOLOGY CO., LTD.|JINGDONG CITY (BEIJING) DIGITS TECHNOLOGY CO., LTD.]]


[[Category:JINGDONG CITY (BEIJING) DIGITS TECHNOLOGY CO., LTD.]]

==Inventor(s)==

[[:Category:Xinzuo Wang of Beijing (CN)|Xinzuo Wang of Beijing (CN)]][[Category:Xinzuo Wang of Beijing (CN)]]

[[:Category:Yang Liu of Beijing (CN)|Yang Liu of Beijing (CN)]][[Category:Yang Liu of Beijing (CN)]]

[[:Category:Junbo Zhang of Beijing (CN)|Junbo Zhang of Beijing (CN)]][[Category:Junbo Zhang of Beijing (CN)]]

[[:Category:Yu Zheng of Beijing (CN)|Yu Zheng of Beijing (CN)]][[Category:Yu Zheng of Beijing (CN)]]

==METHOD AND APPARATUS FOR MODEL TRAINING AND DATA ENHANCEMENT, ELECTRONIC DEVICE AND STORAGE MEDIUM - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240037408 titled 'METHOD AND APPARATUS FOR MODEL TRAINING AND DATA ENHANCEMENT, ELECTRONIC DEVICE AND STORAGE MEDIUM

==Simplified Explanation==

The disclosed patent application describes a method and apparatus for model training and data enhancement using a generative adversarial network (GAN) model. The GAN model consists of a generator and two discriminators. The generator generates reference sample data, which is then used as input for the two discriminators. The first discriminator calculates the distance between the reference sample data and preset negative sample data, while the second discriminator calculates the distance between the negative class data (composed of the reference sample data and preset negative sample data) and preset positive sample data. An objective function is determined based on these distances, and the GAN model is trained using this objective function until convergence is achieved.

* The patent application describes a method and apparatus for training a generative adversarial network (GAN) model.
* The GAN model includes a generator and two discriminators.
* The generator generates reference sample data, which is used as input for the discriminators.
* The first discriminator calculates the distance between the reference sample data and preset negative sample data.
* The second discriminator calculates the distance between the negative class data (composed of the reference sample data and preset negative sample data) and preset positive sample data.
* An objective function is determined based on these distances.
* The GAN model is trained using the objective function until convergence is achieved.

Potential Applications:
* This technology can be applied in various fields where data enhancement and model training are required, such as image and video processing, natural language processing, and data synthesis.
* It can be used to generate realistic synthetic data for training machine learning models, reducing the need for large labeled datasets.
* The method can be used to improve the performance of existing models by enhancing the quality and diversity of the training data.

Problems Solved:
* The technology addresses the problem of limited labeled training data by generating synthetic data that closely resembles real data.
* It solves the challenge of data imbalance by generating additional negative samples and balancing the dataset.
* The method helps in improving the performance and generalization ability of machine learning models by enhancing the quality and diversity of the training data.

Benefits:
* The technology enables more efficient and effective model training by generating synthetic data that closely resembles real data.
* It reduces the reliance on large labeled datasets, making it easier and cheaper to train machine learning models.
* The method enhances the performance and generalization ability of models by providing a more diverse and balanced training dataset.


==Original Abstract Submitted==

disclosed are a method and an apparatus for model training and data enhancement, an electronic device and a storage medium. a generative adversarial network model includes a generator and two discriminators, an output of the generator is used as an input of the two discriminators, the method including: generating, by the generator, reference sample data; calculating, by the first discriminator, a first distance between the reference sample data and preset negative sample data; calculating, by the second discriminator, a second distance between negative class data composed of the reference sample data and the preset negative sample data and preset positive sample data; determining an objective function based on the first distance and the second distance; and training the generative adversarial network model by using the objective function until the generative adversarial network model converges, to obtain the generative adversarial network model.



[[Category:G06N3/094]]
[[Category:G06N3/045]]