Based on healthcare big data (e.g. DNA sequencing, personal health, hospital and patient records) and with the assistance of the latest machine learning technologies (e.g. generative adversarial network, self-supervised learning, deep reinforcement learning, and transfer learning), AIR's endeavors in AI+Healthcare revolve around such areas as genetic engineering, biochemistry, medicine, and medical diagnosis, including:Building a robust pre-training model (including both generative and predictive models), to assist the scientific computing, discovery and exploration in related areas;Developing natural language processing and understanding technologies tailored to healthcare and medicine industries, which serves as a knowledge engine capable of reasoning. For technological application, we plan to combine clinical diagnosis and research to generate large numbers of high-quality marked samples and data that will create a closed loop and feedback for machine learning algorithms and systems in related areas, thus accelerating the applied research and development of AI in personal health, population health, smart healthcare, new drug development, and DNA sequencing.
