Building Personal model for Health Navigation
Population based healthcare models, while helpful in treating diseases caused by external stimulus, have proven not so effective for chronic diseases where the functioning of our biological systems degrades over time. We need to collect high-resolution longitudinal data about the individual to model and predict the transitions of their individual health state over time. Event mining allows us to model the relationships between different events that occur during the course of an individual’s life. We can find the impact of different events on the users’ health by initializing the users’ model (personal model) from domain knowledge and model the nature of the relationships using the data collected in their personal logs (personicle).
Event mining can also be used to find previously unknown relationships by combining it with causal discovery algorithms.
– Laleh Jalali, Ramesh Jain, Bringing Deep Causality to Multimedia Data Streams. ACM Multimedia 2015: 221-230
– N. Nag and R. Jain, A navigational approach to health: Actionable guid¬ance for improved quality of life. IEEE Computer, vol. 52, no. 4, pp. 12–20. April 2019.
– Nitish Nag, Vaibhav Pandey, Preston J. Putzel, Hari Bhimaraju, Srikanth Krishnan, Ramesh Jain, Cross-Modal Health State Estimation. ACM Multimedia 2018: 1993-2002