Artificial Intelligence/Machine Learning (AI/ML) performance is predicated on training with good quality data.
However, such data is often difficult to acquire due to ethical concerns, logistic problems, high cost, data bias, and inherent poor data quality.
Privacy restrictions and data regulations further compound the problem of data acquisition, restricting many organisations long-term access to valuable historical data.
Ultimately, this creates the problem of incomplete or biased data which degrade the overall performance of trained AI/ML models.
This technology offer is a controlled synthetic data generation with differential privacy capability for structured (tabular) data.
Its synthetic data engine utilizes conditional GANs (cGANs) coupled with optional differential privacy to synthesize data with similar properties as real data without the associated privacy risks.
The core technology is a synthetic data engine that learns the distribution of the input data and selects the column to generate based on this distribution. Gaussian noise is further added to the gradients to protect the privacy of the data.
The technology can generate data quickly: 10,000 rows, 8 columns in 8 minutes (evaluated on Nvidia GTX1080) and is mainly intended to generate synthetic datasets to address data scarcity, data privacy, and data augmentation. This generative process involves the following features:
This technology can be used for the following types of structured data:
It can be applied in the following use-cases:
Increase the size of your datasets without wasting time to procure new data
Extrapolate known data to generate unavailable or unknown data points
De-bias or equalize the distribution of datasets
Generate rich data, including infrequent scenarios
This synthetic data generation with differential privacy technology provides accessible privacy by design - adding privacy-preserving techniques before, during or after AI training, together with the following benefits:
The technology owner is looking to collaborate with technology partners in the field of AI/ML to co-develop new products/services, and for collaborators to test-bed in pilot projects.