# Promoting Simple Agents: Ensemble Methods for Event-Log Prediction

arXiv: https://arxiv.org/abs/2604.21629 · HTML: https://mlsanity.com/paper/2604.21629/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 43 references that are in the map.
- First posted on arXiv 23 Apr 2026.

## Builds on (most cited first, 12 of 12)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [On the Practical Computational Power of Finite Precision RNNs for Language Recognition](https://mlsanity.com/paper/1805.04908/) (2018)
- [ProcessTransformer: Predictive Business Process Monitoring with Transformer Network](https://mlsanity.com/paper/2104.00721/) (2021)
- [Time and Activity Sequence Prediction of Business Process Instances](https://mlsanity.com/paper/1602.07566/) (2016)
- [PGTNet: A Process Graph Transformer Network for Remaining Time Prediction of Business Process Instances](https://mlsanity.com/paper/2404.06267/) (2024)
- [Embedding Graph Convolutional Networks in Recurrent Neural Networks for Predictive Monitoring](https://mlsanity.com/paper/2112.09641/) (2021)
- [Text-Aware Predictive Monitoring of Business Processes](https://mlsanity.com/paper/2104.09962/) (2021)
- [Event Stream-Based Process Discovery using Abstract Representations](https://mlsanity.com/paper/1704.08101/) (2017)
- [Directly Follows Graphs Go Predictive Process Monitoring With Graph Neural Networks](https://mlsanity.com/paper/2503.03197/) (2025)
- [Time-Aware and Transition-Semantic Graph Neural Networks for Interpretable Predictive Business Process Monitoring](https://mlsanity.com/paper/2508.09527/) (2025)

## Abstract (arXiv)

We compare lightweight automata-based models (n-grams) with neural architectures (LSTM, Transformer) for next-activity prediction in streaming event logs. Experiments on synthetic patterns and five real-world process mining datasets show that n-grams with appropriate context windows achieve comparable accuracy to neural models while requiring substantially fewer resources. Unlike windowed neural architectures, which show unstable performance patterns, n-grams provide stable and consistent accuracy. While we demonstrate that classical ensemble methods like voting improve n-gram performance, they require running many agents in parallel during inference, increasing memory consumption and latency. We propose an ensemble method, the promotion algorithm, that dynamically selects between two active models during inference, reducing overhead compared to classical voting schemes. On real-world datasets, these ensembles match or exceed the accuracy of non-windowed neural models with lower computational cost.