The invention discloses an intelligent early warning method for forcibly isolating the risk tendency of drug rehabilitation personnel. The method mainly comprises the following steps: acquiring a large amount of compulsory isolation drug addiction personnel data of each business system and security system of a drug addiction; a semi-supervised classification algorithm of a Laplace support vector machine is used for mathematical modeling, the behavior characteristics of the forced isolation drug rehabilitation personnel are automatically analyzed, whether the forced isolation drug rehabilitation personnel have the risk tendency is effectively judged, and the risk tendency type of the forced isolation drug rehabilitation personnel with the risk tendency is judged. According to the invention,a large amount of data in the existing service system and security system of drug rehabilitation is fully utilized, the cost is low, the feasibility is high, and the risk tendency of forcibly isolated drug rehabilitation personnel is analyzed, discriminated and early warned. A high-precision model is trained by using a relatively small amount of sample data, potential behavior characteristics offorcibly isolated rehabilitation personnel are automatically predicted, the influence caused by misjudgment is small, and the problem of machine learning under small samples is solved.