Time series causal relationships discovery through feature importance and ensemble models

We tested our algorithms in different datasets to check causal relationships between time series. In “Climate” section, we show the results of applying causality tests to weather-related time series. “Synthetic oil production field” section provides the results of applying causality in time series related to a synthetic reservoir. Finally, in “Actual oil production field” section, we present results of applying causality to a real production dataset from a Pre-Salt oil field.

Some data pre-processing is required before used it as input in the causal pipeline:

• …