Causal Modeling using software called TETRAD V

First, we have to understand: which aspect of Data Science I am addressing.

Data Security and Data Transparency are important aspects of Data Science, and I am addressing those. Secondly, why do we need Causal Modeling, that involves the finding of proper causes and effects of an event?  We need it because statistical correlation is not adequate in every aspect of data security. The differences between the two approaches will be discussed.

Carnegie Mellon University in Pittsburgh has produced some remarkable high quality software called TETRAD V, for handling the modeling of causation. Using that, one can show important results, for example, how to figure out when the data access is insecure.

It allows one to use the notion of causal graphs, to manipulate such graphs, to generate patterns from them, and to compare graphs. One can build Bayes parametric models or Structural Equation Models (SEM) parametric models. These concepts will be explained.

There are various other components, such as Knowledge box, Calculate Box, Estimation Box, etc. It also uses various forms of Search techniques. One of the most important usages of TETRAD is the proper manipulation of raw data, in the cases where no previous data (that could have yielded “expected value”) are available.