The 36th International Conference on Machine Learning (ICML 2019), the top ML conference in the world, will be held in Long Beach, CA, from Sunday, June 9th, to Saturday, June 15th.
doc.ai, a Diamond Sponsor of the conference, will present Federated Learning technology for healthcare along with other core technologies we build to advance medical research at large.
From day one our values have been ensuring the privacy and security of our users’ data since we work with the most sensitive data – health data. As a result, doc.ai is pioneering Federated Learning in healthcare which inherently brings privacy and security. We’re excited to share developments with you.
If you are attending ICML 2019, we hope you’ll come over to our booth #217 to chat with our research team and learn more about the projects at doc.ai that will give you the freedom to build impactful solutions to transform healthcare.
For those of you unfamiliar with ICML, the conference highlights the latest, state-of-the-art research in domains such as deep learning, self-attention mechanisms, meta-learning, one-shot learning, generative models, optimization methods, reinforcement learning, imitation learning, adversarial machine learning, and more.
Below is a sneak peak at the core technology that will be displayed at the doc.ai booth:
Phenomenal face: medical selfie predictions.Pupil: medication and supplement predictions.Flea: federated learning, technology that allows decentralized ML training without direct access to the data.Imagine API: Python3 SDK for medical predictions (you can use our predictive models through Jupyter Notebook).TensorIO, a declarative on-device ML framework. If you are interested in running and training models on mobile phones, we encourage you to try out this open-sourced framework.Genome browser, a genome features exploration tool.
You can learn more about our projects from the talks and demos of the doc.ai team and advisors on Expo Day, Sunday, June 9th.
doc.ai talks (Hall B):
2pm: “Mapping high-throughput DNA sequencing reads with learned index structures” by Mike Lin, genome scientist, former MIT, doc.ai advisor.
3pm: “Visualizing personal genomics with genewall-tensor” by Win Van Criekinge, Professor in Computational Genomics and Bioinformatics at Ghent University, doc.ai advisor.
4:30pm: “Flea: An open source framework for federated learning” by Neeraj Kashyap, Head of AI at doc.ai.
5:30pm: “Terabyte-scale exposomic, genetic, sociodemographic, and health data for building your cohort for precision medicine” by Walter de Brouwer, doc.ai CEO; Chirag Patel and Arjun Manrai, associate professors’ at Harvard Medical School and doc.ai advisors.
You can learn more about the talks and look at the schedule here.
Transcending Federated Learning through MannaDeclarative On-Device Machine LearningImagine-api, a Python3 SDK for performing predictions using doc.ai TensorFlow models
Check out our open-source projects here and see you soon in Long Beach!