They prefer to fit data to variants of a “logistic” curve.

它们倾向于用数据去迎合“逻辑”曲线上的变量。