17947778. USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR simplified abstract (Capital One Services, LLC)
Contents
- 1 USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Original Abstract Submitted
USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR
Organization Name
Inventor(s)
Austin Walters of Savoy IL (US)
Galen Rafferty of Mahomet IL (US)
Jeremy Goodsitt of Champaign IL (US)
Anh Truong of Champaign IL (US)
USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR - A simplified explanation of the abstract
This abstract first appeared for US patent application 17947778 titled 'USING A TUNABLE PRE-TRAINED DISCRIMINATOR TO TRAIN A GENERATOR AND AN UNTRAINED DISCRIMINATOR
Simplified Explanation
The patent application describes a system that implements a tunable pre-trained discriminator in a machine learning model, such as a general adversarial network. The system generates training data using a generator and sends it to a first pre-trained discriminator and a second untrained discriminator. The system receives sets of labels from both discriminators, selects a label, and provides the selected label and corresponding data records to further train the generator.
- The system implements a tunable pre-trained discriminator in a machine learning model.
- Training data is generated using a generator and sent to a pre-trained and untrained discriminator.
- Labels are received from both discriminators, a label is selected, and the generator is further trained with the selected label and data records.
Potential Applications
The technology described in the patent application could be applied in various fields such as image generation, natural language processing, and anomaly detection.
Problems Solved
This technology helps improve the performance and efficiency of machine learning models by utilizing a tunable pre-trained discriminator to provide better training data for the generator.
Benefits
The benefits of this technology include enhanced training of machine learning models, improved accuracy in generating data, and increased overall performance of the system.
Potential Commercial Applications
One potential commercial application of this technology could be in the development of advanced AI systems for industries such as healthcare, finance, and cybersecurity.
Possible Prior Art
Prior art in the field of machine learning and generative models may include research papers, patents, and existing systems that utilize pre-trained discriminators in machine learning models.
Unanswered Questions
How does this technology compare to existing methods in terms of training efficiency and model performance?
This article does not provide a direct comparison between this technology and existing methods in terms of training efficiency and model performance. Further research or testing may be needed to evaluate the effectiveness of this technology in comparison to other methods.
What are the potential limitations or challenges in implementing this technology in real-world applications?
The article does not address potential limitations or challenges in implementing this technology in real-world applications. Factors such as computational resources, data quality, and model scalability could be important considerations when deploying this technology.
Original Abstract Submitted
Systems as described herein may implement a tunable pre-trained discriminator in a machine learning model, such as a general adversarial network. A server may generate training data using a generator of the machine learning model. The server may send the training data to a first discriminator (e.g., a pre-trained discriminator) and a second discriminator (e.g., an untrained discriminator). The server may receive a first set and a second set of labels from the first discriminator and the second discriminator, respectively. The server may select a label from either the first or the second set of labels. Accordingly, the server may provide the selected labels and the corresponding data records to further train the generator of the machine learning model.