DIFF-ERO: A Conformance-Aware Loss for Deep Learning in Process Mining

2026-06-12

AI Summary

This work addresses the limitation of existing deep learning approaches in process mining, which typically optimize only local next-step predictions and struggle to ensure global behavioral consistency with the underlying process control-flow structure. To overcome this, the authors propose DIFF-ERO, a differentiable entropy-based stochastic conformance loss that, for the first time, directly embeds structural conformance signals into the training process. By explicitly modeling control flow through batch-level soft edge transition matrices and jointly optimizing them with cross-entropy loss, DIFF-ERO guides the model to internalize the true process structure. The method is architecture-agnostic and consistently improves prediction performance in structure-sensitive scenarios across multiple benchmarks, yielding learned stochastic automata that more closely approximate the ground-truth process models, thereby demonstrating its effectiveness.

Abstract

Deep learning has driven many recent advances in process analytics, especially for predictive and prescriptive monitoring. However, standard objectives such as cross-entropy optimize local next-step likelihoods and only implicitly capture control-flow structure. As a result, models can achieve high token-level accuracy while permitting imprecise global behaviour. We introduce DIFF-ERO, a conformance-aware loss function for deep learning models on process data. DIFF-ERO is a differentiable formulation of entropy-based stochastic conformance that incorporates control-flow information during training. Our approach constructs batch-level stochastic transition matrices with soft edge memberships, allowing structural precision and recall signals to directly inform backpropagation. The loss is model-agnostic and can be applied whenever the final representation parametrizes stochastic transitions. We instantiate DIFF-ERO in transformer encoder-decoder pipelines for next-activity prediction and use it jointly with cross-entropy to analyse its theoretical components with respect to convergence. Across benchmarks comparing other loss functions and targets, DIFF-ERO shows improved predictive performance where structure matters most while maintaining parity elsewhere. At the same time, the learned stochastic automaton converges towards the structural ground truth, indicating that the network internalizes process model structure.

Problem

• process mining
• deep learning
• conformance
• control-flow
• loss function

Innovation

• DIFF-ERO
• conformance-aware loss
• stochastic conformance
• differentiable process mining
• control-flow structure