In this seminar talk, we discuss methods for statistical hypothesis generation, such as point estimation, regression analysis and principal component analysis, and generalize them into a cohesive unified method for statistical hypothesis generation. From the unified method, the aforementioned methods appear as special cases under supplementary assumptions, which  clarifies precisely their similarities and differences. As the unified method employs the scientific ideas originally articulated by Jacob Bernoulli, C. F. Gauss, Karl Pearson et alii, the talk will also summarize the philosophical tenets of statistical hypothesis generation.