Temporal Probabilistic Asymmetric Multi-Task Learning model is a probabilistic asymmetric multi-task learning framework that allows asymmetric knowledge transfer between tasks at different timesteps, based on the uncertainty. While existing asymmetric multi-task learning methods consider asymmetric relationships between tasks as fixed, the task relationship may change at different timesteps in time-series data. Moreover, knowledge obtained for a task at a specific timestep could be useful for other tasks in later timesteps. This model is trained on clinical time-series prediction tasks on four datasets, on which it shows strong performance over the baseline symmetric and asymmetric multitask learning models, without any sign of negative transfer. Several case studies with learned knowledge graphs show that this model is interpretable, providing useful and reliable information on model predictions. This interpretability of the model will be useful in building a safe time-series analysis system for large-scale settings where both the number of time-series data instances and timestep are extremely large, such that manual analysis is impractical.
Input variables : EHR data
Output Variables : Predictions for clinical task like fever, infection, length of stay and mortality
| Statistical | : | Mallow's CP | R Squared | Mean Square Error | Adjusted R Squared | Mean Absolute Error | Huber Loss |
| Business | : | Population at High Risk of Disease | Population at High Risk of Disease | Risk by Geography | Risk by Geography | Risk by Demographics | Risk by Demographics | Risk by Clinical Parameters | Risk by Clinical Parameters | Optimized Hospital Resource Utilization | Decreased Cost of Care | Decreased Patient Visits |
| Infrastructure | : | Log Bytes | Logging/User/IAMPolicy | Logging/User/VPN | CPU Utilization | Memory Usage | Error Count | Prediction Count | Prediction Latencies | Private Endpoint Prediction Latencies | Private Endpoint Response Count |
Visit Model : github.com
Additional links : arxiv.org
| Model Category | : | Public |
| Date Published | : | February, 2021 |
| Healthcare Domain | : | Provider |
| Code | : | github.com |
| Health Risk Management |
| Health Risk Prediction |
| Risk Progression |