Multi-task learning shows promise for predictive process monitoring

A new study explores the potential of multi-task learning (MTL) for Predictive Process Monitoring (PPM), a field that forecasts the unfolding of organizational processes. While deep learning has advanced PPM, most existing methods use single-task learning (STL), which is inefficient and misses potential synergies. This research presents the first comprehensive empirical study of MTL for PPM, evaluating various task combinations, neural architectures, and optimization methods. The findings indicate that MTL offers substantial improvements in next-activity prediction and effectively mitigates class imbalance, with task balancing being particularly crucial for lower-capacity models. AI

IMPACT This research could lead to more efficient and accurate predictive process monitoring systems by leveraging multi-task learning.

RANK_REASON The cluster contains a research paper detailing a new methodology for an existing field. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• arXiv
• Class Imbalance Invisibility
• deep learning
• multi-task learning
• Neural architectures for adaptive behavior
• Optimization Methods and Software
• Predictive Process Monitoring
• single-task learning

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →