In this paper‎, ‎ an adaptive technique for the prediction of binary response variable by combining fuzzy least squares method with statistical logistic regression was proposed‎. ‎Also for each α-cut‎, ‎ bootstrap technique was used to discuss the problem of statistical inference when there was crisp inputs and fuzzy outputs‎. ‎A simulation study and a numerical example in the clinical field is provided to check the efficiency of the proposed approach‎.

Keywords: Regression analysis, fuzzy, logistic, bootstrap