We have presented a novel application of denoising autoencoders to detect anomalies in business process data. Our approach does not rely on any prior knowledge about the process itself. Also, we do not rely on a clean dataset for the training; our...
Credentials & highlights
- Project Views