filmov
tv
“Learn Everything, Know Nothing: The Medical Applications of Privacy-Preserving AI”

Показать описание
2020 Faculty of Science Undergraduate 3MT competition
1st Place Winner: Physical, Computation and Data Science category
Delica Leboe-McGowan – “Learn Everything, Know Nothing: The Medical Applications of Privacy-Preserving AI”
Supervisor: Dr. Noman Mohammed, Department of Computer Science, University of Manitoba
Delica Leboe-McGowan’s project uses a method called secret sharing to ensure that patients’ genomic data can remain private while it is used to train artificial intelligence (AI) models for cancer detection. Without ever compromising the security of patients’ private information, Delica’s neural networks find patterns in their genomic profiles that allow her AI model to diagnose breast cancer with over 90% accuracy after less than a minute of training.
1st Place Winner: Physical, Computation and Data Science category
Delica Leboe-McGowan – “Learn Everything, Know Nothing: The Medical Applications of Privacy-Preserving AI”
Supervisor: Dr. Noman Mohammed, Department of Computer Science, University of Manitoba
Delica Leboe-McGowan’s project uses a method called secret sharing to ensure that patients’ genomic data can remain private while it is used to train artificial intelligence (AI) models for cancer detection. Without ever compromising the security of patients’ private information, Delica’s neural networks find patterns in their genomic profiles that allow her AI model to diagnose breast cancer with over 90% accuracy after less than a minute of training.