Average treatment effect for paired data

A computer program product comprising:

• one or more computer readable storage media and program instructions stored on the one or more computer readable storage media, the program instructions comprising:
• program instructions to identify a plurality of data variables within a stream of a multivariate event dataset;
• program instructions to formalize a causal inference between a first identified data variable associated with historical occurrences and a second identified data variable associated with a different occurrence within the multivariate event dataset by determining a causal connection strength for the first identified data variable and the second identified data variable;
• program instructions to train a data model with the multivariate event dataset based on the causal inference, wherein the data model estimates a plurality of treatment effects between the first identified data variable and the second identified data variable;
• program instructions to calculate an inverse propensity score for the data model based on a type of identified variable, a predetermined time associated with the identified variable, and the causal connection strength between the first identified data variable and the second identified data variable;
• program instructions to validate the inverse propensity score by splitting the multivariate event dataset into a training data set based on an identified optimal time window;
• program instructions to, responsive to the inverse propensity score being validated, optimize the data model by strengthening the causal inference between the first identified data variable and the second identified data variable;
• and
• program instructions to, responsive to the inverse propensity score failing validation, optimize the data model by removing the causal inference between the first identified data variable and the second identified data variable.

A computer system comprising:

• one or more computer processors;
• one or more computer readable storage media;
• and
• program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more computer processors, the program instructions comprising:
• program instructions to identify a plurality of data variables within a stream of a multivariate event dataset;
• program instructions to formalize a causal inference between a first identified data variable associated with historical occurrences and a second identified data variable associated with a different occurrence within the multivariate event dataset by determining a causal connection strength for the first identified data variable and the second identified data variable;
• program instructions to train a data model with the multivariate event dataset based on the causal inference, wherein the data model estimates a plurality of treatment effects between the first identified data variable and the second identified data variable;
• program instructions to calculate an inverse propensity score for the data model based on a type of identified variable, a predetermined time associated with the identified variable, and the causal connection strength between the first identified data variable and the second identified data variable
• program instructions to validate the inverse propensity score by splitting the multivariate event dataset into a training data set based on an identified optimal time window;
• program instructions to, responsive to the inverse propensity score being validated, optimize the data model by strengthening the causal inference between the first identified data variable and the second identified data variable;
• and
• program instructions to, responsive to the inverse propensity score failing validation, optimize the data model by removing the causal inference between the first identified data variable and the second identified data variable.