Systems and methods for modeling and analyzing networks

A computer-implemented method of analyzing a dataset having two or more variables, comprising:

• (i) receiving, at a network inference engine, under control of a first processor, a dataset having two or more variables;
• (ii) inferring, by the first processor, at the network inference engine, one or more global causal network computer models representative of the dataset, and each of such global causal network computer models containing equations describing the relationships between said variables, by:
• generating, at a network fragment builder, a plurality of local computer models, each local computer model describing connections between two or more of the variables in the dataset, and
• combining a plurality of the local models to generate the one or more global causal network computer models, such that the one or more global causal network computer models include two or more nodes corresponding to the two or more variables in the dataset;
• and
• (iii) simulating said one or more global causal network computer models, at a forward simulation engine, under control of a second processor, to predict the impact of the change made to the value of one or more first variables on the values of one or more second variables.