3results about How to "Reasonable value" patented technology

The application provides a similarity feature constraint-based inverse synthesis prediction method, medium and equipment; wherein the method is: converting a to-be-predicted product into a corresponding product SMILES sequence; using a graph neural network to extract graph structure features corresponding to all characters in the product SMILES sequence; fusing the graph structure features and word vector representations to obtain a fusion vector; inputting the fusion vector matrix into a Transformer encoder and decoder to obtain a reactant set SMILES sequence, and converting the reactant set SMILES sequence to obtain corresponding reactants. The method integrates the graph structure features of the product molecule into the prediction features, improves the prediction accuracy effect, learns the commonality between the reactant and product sets through feature constraint, uses the commonality features to better help the Transformer model to perform inverse synthesis prediction, and further improves the prediction accuracy of the model.

The application discloses a network protocol fuzzy test seed evaluation method and device based on multi-standard combination weighting. In order to realize multi-standard combination weighting, first, evaluation indexes for measuring the value of a network protocol fuzzy test seed are selected, including basic evaluation indexes, future potential indexes and network protocol related indexes. Then, pre-fuzzy testing is performed to obtain item evaluation index values of sample seeds. Based on the evaluation index values, an entropy weight method is used to determine first weights of all evaluation indexes, and a maximum deviation method is used to determine second weights of all evaluation indexes. The evaluation weights of the evaluation indexes are determined by comprehensively considering the first weights and the second weights. According to the evaluation weights, all evaluation index values of the sample seeds are integrated to obtain evaluation values of the sample seeds. The application can improve the network protocol information utilization rate, thereby ensuring the stability of the seed selection process, improving the overall quality of the selected seeds, and further improving the efficiency of the fuzzy test and the vulnerability discovery capability.

The invention provides a method for correcting a capacitance measurement value under large electric leakage, and provides a parallel binary model for measuring the capacitance value of a gate oxide layer of a gate oxide MOS (Metal Oxide Semiconductor) device, and the parallel binary model is a parallel circuit model formed by connecting a parallel capacitor Cp and an equivalent resistor Rp in parallel; selecting the parallel binary model through an LCP tester, and measuring the gate oxide MOS device by adopting a frequency f1 to obtain a parallel capacitance Cp1 and a parallel conductance Gp1; calculating a dissipation factor D1 when D1lt; when the value is 0.1, the measured shunt capacitance Cp1 is the capacitance value of the gate oxide layer in the gate oxide MOS device; when D1gt; when the capacitance value is 0.1, a parallel binary model is selected through an LCP tester, the gate oxide MOS device is measured by adopting the frequency f2, parallel capacitance Cp2 and parallel conductance Gp2 are obtained, and the actual capacitance value C of the gate oxide layer of the gate oxide MOS device is obtained.