A clinical examination allows clinicians to first make a qualitative diagnosis. However, in clinical practice, the same disease is often classified into different stages according to the degree to which its symptoms threaten the patient’s health, progress of the disease, or difficulty of its treatment. The severity of such diseases is quantitatively assessed with interval or ordinal scales in the form of biomarkers known to be closely related to the disease or scores given in questionnaires or check sheets. In other cases, as in the concept of “me-byo,” it may be more appropriate to view healthy and sick individuals as a continuum of change rather than to divide them into groups with definite diagnostic criteria. In such cases, a model that estimates the severity of the disease rather than merely the presence or absence of the disease would be more useful as it aims to predict existing quantitative indicators rather than discriminate between two or multiple groups.