MTMA: Multi-task multi-attribute learning for the prediction of adverse drug–drug interaction

作者：

Highlights：

• A novel learning model MTMA is proposed for ADDI prediction.

• Two interpretable tensors are designed to uncover the adverse mechanisms among drugs.

• l2,1-norm regularization is adopted to explore the leading factors of ADDIs.

• Experimental results demonstrate the effectiveness of MTMA in ADDI prediction.

摘要

•A novel learning model MTMA is proposed for ADDI prediction.•Two interpretable tensors are designed to uncover the adverse mechanisms among drugs.•l2,1-norm regularization is adopted to explore the leading factors of ADDIs.•Experimental results demonstrate the effectiveness of MTMA in ADDI prediction.

论文关键词：Adverse drug–drug interaction,Multi-task,Multi-attribute,Supervised learning,Tensor decomposition

论文评审过程：Received 27 October 2019, Revised 24 April 2020, Accepted 26 April 2020, Available online 30 April 2020, Version of Record 5 May 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.105978