20230044378. SYSTEMS AND METHODS FOR ARCHITECTURE EMBEDDINGS FOR EFFICIENT DYNAMIC SYNTHETIC DATA GENERATION simplified abstract (Capital One Services, LLC)
Contents
SYSTEMS AND METHODS FOR ARCHITECTURE EMBEDDINGS FOR EFFICIENT DYNAMIC SYNTHETIC DATA GENERATION
Organization Name
Inventor(s)
Austin Walters of Savoy IL (US)
Jeremy Goodsitt of Champaign IL (US)
Anh Truong of Champaign IL (US)
Galen Rafferty of Mahomet IL (US)
Grant Eden of San Francisco CA (US)
SYSTEMS AND METHODS FOR ARCHITECTURE EMBEDDINGS FOR EFFICIENT DYNAMIC SYNTHETIC DATA GENERATION - A simplified explanation of the abstract
This abstract first appeared for US patent application 20230044378 titled 'SYSTEMS AND METHODS FOR ARCHITECTURE EMBEDDINGS FOR EFFICIENT DYNAMIC SYNTHETIC DATA GENERATION
Simplified Explanation
The patent application describes systems and methods for efficiently generating synthetic data using architecture embeddings. Here are the key points:
- The system retrieves a set of rules associated with a specific data profile.
- It generates multiple hyperparameter sets for generative adversarial networks (GANs) that satisfy the rules.
- The system creates mappings between the hyperparameter sets and the data profile, storing them in a hyperparameter library.
- When a request for synthetic data is received with a different data profile, the system selects a hyperparameter set from the library that matches the profile.
- It builds a GAN using the selected hyperparameter set and generates a synthetic data set using the GAN.
Potential applications of this technology:
- Data augmentation: Synthetic data can be used to increase the size and diversity of training datasets for machine learning models.
- Privacy protection: Synthetic data can be used as a substitute for sensitive or confidential data, reducing the risk of data breaches.
- Testing and simulation: Synthetic data can be used to simulate various scenarios and test the performance of systems or algorithms.
Problems solved by this technology:
- Efficient data generation: The system automates the process of generating synthetic data by using architecture embeddings and hyperparameter sets, reducing the time and effort required.
- Rule-based data generation: The system ensures that the generated data adheres to specific rules or constraints defined by the data profile, improving the quality and relevance of the synthetic data.
Benefits of this technology:
- Scalability: The system can generate large amounts of synthetic data quickly, allowing for the creation of diverse datasets for various applications.
- Flexibility: By using different hyperparameter sets, the system can generate synthetic data that closely matches different data profiles, providing flexibility in data generation.
- Privacy preservation: Synthetic data can be used as a privacy-preserving alternative to real data, protecting sensitive information while still enabling analysis and model training.
Original Abstract Submitted
systems and methods for architecture embeddings for efficient dynamic synthetic data generation are disclosed. the disclosed systems and methods may include a system for generating synthetic data configured to perform operations. the operations may include retrieving a set of rules associated with a first data profile and generating, by executing a hyperparameter search, a plurality of hyperparameter sets for generative adversarial networks (gans) that satisfy the set of rules. the operations may include generating mappings between the hyperparameter sets and the first data profile and storing the mappings in a hyperparameter library. the operations may include receiving a request for synthetic data, the request indicating a second data profile and selecting, from the mappings in the hyperparameter library, a hyperparameter set mapped to the second data profile. the operations may include building a gan using the selected hyperparameter set and generating, using the gan, a synthetic data set.
