Causal Discovery

Understanding cause and effect based on observational data primarily means adjusting for potential “confounders” (indirect explanations for a target event). Missing information therefore leads to flawed conclusions on causal effects – and this also holds if you can analyze just parts of a complex object at a time. Xplain Data’s holistic Object Analytics therefore constitutes novel opportunities to uncover potential cause and effect relationships: the wealth of information stored in such an object model is used to quickly evaluate millions of potential confounders, and present only those factors the effect of which cannot be explained “via other factors” – still no proof without experiment – but we can segregate away myriads of meaningless correlations and help you to quickly get to relevant hypothesis on causal effects.

Knowing causal dependencies means being able to influence a system – a major step towards intelligent systems in real world environments.

Example: Factors potentially causing breast cancer, including a graphical representation which visualizes direct and indirect effects on the target of analysis.

The Causal Discovery algorithms are embedded into the Object Explorer and its interactive usability concept: Once configured, just click at a target, and you will see what drives or potentially causes this target. The domain expert may reject certain factors, ask for alternatives … such finally arriving at results, which combine knowledge from data with domain expertise.

Any of those algorithms can also be used from within, e. g., a Python environment. Listen to Paula to understand our Causal Discovery approach on an intuitive level.