The article discusses a method to figure out how many hidden factors are influencing a set of observed variables, even when we don't know the exact relationships between them. By using a special type of statistical model, the researchers were able to identify the hidden factors and the connections between the observed variables. They developed a mathematical program that can accurately estimate the number of hidden factors and the structure of the relationships among the observed variables. This method is useful for analyzing large amounts of data where the number of variables is high compared to the number of samples.