DIFF-ERO: A Conformance-Aware Loss for Deep Learning in Process Mining
For process mining practitioners, this addresses the problem of deep learning models achieving high token accuracy but imprecise global behavior, though the improvement is incremental over existing methods.
The authors introduce DIFF-ERO, a differentiable conformance-aware loss for deep learning in process mining that incorporates control-flow structure during training. It improves predictive performance on structure-sensitive tasks while maintaining parity elsewhere, and the learned automaton converges toward structural ground truth.
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.