Learning causation event conjunction sequences
This work addresses event sequence analysis for domains like healthcare or finance, but it appears incremental as it compares existing methods without introducing a new paradigm.
The paper tackled the problem of learning causation as conjunctions of events in sequences, finding that a histogram-based algorithm significantly outperformed all artificial neural network methods tested.
This is an examination of some methods that learn causations in event sequences. A causation is defined as a conjunction of one or more cause events occurring in an arbitrary order, with possible intervening non-causal events, that lead to an effect. The methods include recurrent and non-recurrent artificial neural networks (ANNs), as well as a histogram-based algorithm. An attention recurrent ANN performed the best of the ANNs, while the histogram algorithm was significantly superior to all the ANNs.