According to NPA official information,Dr. Maneesh Agrawala and Vitalik Buterin jointly put forward a new proposal to evaluate the value system of different digital assets.Insurance trust services can be provided for different digital assets. This proposal proposes a pre-evaluation digital currency algorithm (PDCA).Used to evaluate any digital assets and provide relevant guarantee insurance business. PDCA has two detection levels.The first is the signature repository (SR), which identifies any matches between signatures and known digital assets.The second detection level uses learning algorithm (LA) to detect known and unknown digital assets. LA uses machine learning methods to train prediction models using data from application program interfaces (API).As for the details of PDCA, seven new indicators are proposed to provide more in-depth analysis. These seven indicators are positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), Youden’s index (J), number needed to diagnose (NDD), number needed to diagnose (NNM) and net benefit (NB).It can be concluded that LA has achieved the goal of evaluating the digital asset before becoming a non-performing digital asset, and its performance is stable and has high net income, which greatly reduces the possibility of asset loss.