Abstract
In usual statistical data analysis, we describe statistical data by exact values. However, in modern complex and large-scale systems, it is difficult to treat the systems using only exact data. In other words, it occurs that we have to treat non-exact data involving human vagueness. In this paper, we define these data as fuzzy data and propose new methods which can treat the fuzzy data by usual statistical methods. Concretely, using fuzzy data from q populations consisting of regression models, we develop the methods to get the maximum likelihood estimates of their parameters by usual statistical methods under fuzzy observations. Furthermore, we investigate the validity of our methods by computer simulations under realistic situations.
