Skip to main content
Fig. 2 | BMC Medical Informatics and Decision Making

Fig. 2

From: Developing a novel causal inference algorithm for personalized biomedical causal graph learning using meta machine learning

Fig. 2

An illustration of different learning schema for multi-task causal graph learning. a Decoupled learning methods solve each task independently without considering the similarities among tasks; b Joint learning methods introduce an additional set of parameters to model the shared commonality among all tasks and learn a separate variation parameter for each task; and c Our method applies a new framework based on meta machine learning principle and learns both commonality and variations for each task. Compared with baselines, meta-learning enables the sharing of knowledge across tasks for enhanced prediction performance and the rapid adoption of knowledge to new tasks for efficient causal graph learning

Back to article page